What does the future hold for MRI?
- Figure 1. Neurological exams at 3T can easily achieve better than 0.4mm in-plane resolution in short scan times. The above figure shows clinical 3T neuro scans. Left: T2w. TSE sequence, 0.3mm resolution, 5.2mins Right: Non-contrast TOF angiogram, 0.25mm resolution, 8 mins
Magnetic resonance imaging (MRI) is already the most flexible diagnostic imaging modality, allowing us to characterise many aspects of the living patient from metabolism and physiology to tissue microstucture. Here we consider the future directions for MRI technology and predict that even after 30 years of continuing developments there are still further major advances which can be made in diagnostic MRI.
by Prof. Andrew Blamire
Predicting the future developments in any major technology is always a challenge, and MRI is no exception. In the early days of MRI predictions were made that imaging would not be possible above a field strength of 0.23T , and that the process by which body sections are imaged (slice section) even violated a fundamental principle of physics, namely the Uncertainty Principle, . Fortunately for clinical radiology, such gloomy predictions were not realised and virtually every district hospital in Europe now operates scanners based on high field magnets. Nowadays there is a wealth of worldwide research underpinning the future commercial developments of diagnostic technologies. The future path for MRI can be predicted by identifying the most promising research findings, which over time (typically five to 10 years) and with adoption by scanner manufacturers, will arrive in the clinic - the MRI of the future therefore depends on MRI research today. So what might we expect from MRI in the coming decade?
MR scanner hardware
In the mid-1990s research institutions began ordering high-field scanners operating above the common 1.5 Tesla (T) clinical limit. This was motivated by the basic physics of scanning, which states that signal strength (and hence image quality) increases directly with magnetic field strength. As a result, high field strength MRI (3T) scanners have been available as clinical products for more than five years and are gradually being established as the routine field strength, a trend which is firmly set to continue and is likely to see most hospitals working at 3T in the next decade. Compact scanner design allows the new generation 3T systems to slot into old 1.5T scanner sites. Field strength for neurological examinations translates directly into increased sensitivity (double at 3T vs. 1.5T), which can be used to obtain higher resolution images or traded for speed (shorter examination time). Interestingly, the duration of a typical MRI patient investigation has not declined significantly in the last few years, with faster image acquisition simply allowing a wider range of image types to be collected in the individual. Neurological exams at 3T can easily achieve better than 0.4mm in-plane resolution in short scan times [Figure 1]. Currently, body applications of 3T are more variable in their reproducibility (e.g. cardiac methods) and, while 1.5T remains the current standard, developments in scanner technology will eventually realise the benefit of 3T in these clinical areas too. Although many ultra high-field (7T and above) brain imaging systems are in research use, these are unlikely to have any major impact as a clinical technology in their own right, but as with all cutting edge developments, innovations made at 7T will trickle down and affect routine clinical products. At the other extreme of field strength however, hardware is being developed using a technique known as pre-polarised MR, which can achieve 1.5T image quality in peripheral limbs using only a 0.1T magnet, and provide artefact-free images in the presence of metallic implants – all in a prototype scanner costing around $50,000 . For specialist clinics, the commercial versions of such technology will offer a package of high quality images, patient convenience and affordability.
Another area of hardware development that continues to develop is receiver coil technology. All modern scanners have multiple receiver channels – a development which began in 1990 with the introduction of phased-array spine coils – and this area moves onward almost unabated. Small receiver coils have high local sensitivity but limited field of view (FOV). Combining a number of such small coils into a single device – the multi-channel coil – restores full FOV while maintaining sensitivity. Eight and 16 channel devices are increasingly common and systems have been designed for modular upgrade to 24, 32 channels and beyond. Research systems show benefits of at least 32 channels in brain imaging , and up to 100 channel devices have been designed , but the sheer complexity of such systems means that they are less likely to be clinically relevant.
Integrated modality scanners are a particularly exciting engineering development that will impact on patient assessment, particularly in specialist oncology centres. Positron emission tomography (PET) is indicated for diagnosis in cancer but provides no structural data. PET systems are already integrated with CT systems to provide anatomical detail, but still suffer from poor soft tissue contrast. Patients therefore may undergo separate MRI investigation to collect the soft tissue information. Integrating PET with MRI into a single modality is therefore a prime technological goal. Both rodent and human realisation of combined (simultaneous) PET-MRI scanners have been reported in the last few years, overcoming the challenge of operating the PET detector system within the hostile environment of a high magnetic field . Once fully commercialised, this technology will improve data quality, allow more accurate rendering of metabolic to structural data and improve patient throughput, as PET and MRI can then be collected simultaneously in the same scan session.
- Figure 2. Integrated examination protocols collecting fully quantitative measures are already possible in the brain, providing data on metabolic, physiological, micro- and macro-structural tissue integrity. The above figure shows an integrated quantitative 3T neuro exam. Top left : T1w relaxation map. Top right T2 relaxation map
MRI is different from all other diagnostic modalities in that the hardware configuration plays little role in the image contrasts that can be collected, these instead being dictated by software through the scanning sequences. Several new methods show potential. Perfusion measurements have become central to many neurological protocols (particularly in acute stroke), based around kinetic tracking of a bolus contrast agent injection. For the past 10 years alternative non-contrast agent approaches have been developing, known collectively as arterial spin labelling (ASL), and at 3T these sequences show real promise of providing clinically useful images of cerebral blood flow CBF [Figure 2, bottom right].
Image contrast generated by any MRI sequence relies (partly) upon the rate at which the transient MR signal is decaying away (the relaxation time T2) the time at which the signal is recorded (the echo time, TE). Recently clinical research findings have been published using ultra-short echo time (UTE) sequences where TE has been reduced from its typical range of 1-10 milliseconds down to 10-50 microseconds . This has created a whole new set of image contrasts allowing direct imaging of tissues which have traditionally not been seen by MRI
(e.g. ligaments and tendons, ).
While conventional T1 and T2 weighted scans will always remain the bed-rock of the clinical examination, it is clear that truly quantitative scan types will be needed as we enter the era of personalised medicine, where non-invasive monitoring of treatment response will be increasingly important. Integrated examination protocols collecting fully quantitative measures are already possible in the brain [Figure 2], providing data on metabolic, physiological, micro- and macro-structural tissue integrity, and these may well become the norm. These fundamental biological parameters are inter-related and the relationship between them is likely to provide additional detail on tissue injury or disease. Advanced image processing techniques are currently being developed that will provide the radiologist with tools to assimilate the many types of data.
New contrast agents
As clinical treatment of many conditions become increasingly personalised, improved diagnosis, and hence specific selection of therapeutic treatments based on knowledge of disease phenotype, will become important. For many years PET has offered the ability to map molecular pathways, receptor density, etc, via appropriate radioligands, but PET cannot provide the level of spatial resolution offered by MRI, and is limited in availability. MRI is responding to this challenge through the development of targeted MR contrast media that bind to specific molecular or cellular targets. This work is currently in its infancy, but recent data in animal models have shown the power to detect the earliest phase of brain injury . Although the transfer of such technology into useful clinical compounds requires extensive development and product licensing, these data show clear potential for this technique.
MRI has continued to rapidly develop since its introduction as a clinical tool in the early 1980s. Widespread use of 3T scanners is already becoming a reality and future developments in coil technology and new image contrasts will continue to provide new tools for clinical diagnosis. Combined modalities and targeted contrast media are further on the clinical horizon, but will become a reality for specialist referral centres in the coming years.
1. Hoult DI, Lauterbur PC. Sensitivity of the Zeugmatographic Experiment Involving Human Samples. Journal of Magnetic Resonance 1979; 34: 425-433.
2. Hoult DI. Zeugmatography - Criticism of Concept of a Selective Pulse in Presence of a Field Gradient. Journal of Magnetic Resonance 1977; 26: 165-167.
3. Venook RD et al. Prepolarized magnetic resonance imaging around metal orthopedic implants. Magnetic Resonance in Medicine 2006; 56: 177-186.
4. Wiggins GC et al. 32-Channel 3 tesla receive-only phased-array head coil with soccer-ball element geometry. Magnetic Resonance in Medicine 2006; 56: 216-223.
5. Wiggins GC et al. A 96-channel MRI system with 23- and 90-channel phase array head coils at 1.5 Tesla. Proceedings of the 13th Scientific Meeting of the International Society for Magnetic Resonance in Medicine.
2005. Miami, Florida, USA.
6. Schlemmer HPW et al. Simultaneous MR/PET imaging of the human brain: Feasibility study. Radiology 2008; 248: 1028-1035.
7. Gatehouse PD, Bydder GM. Magnetic resonance imaging of short T-2 components in tissue. Clinical Radiology 2003; 58: 1-19.
8. Robson MD, Bydder GM. Clinical ultrashort echo time imaging of bone and other connective tissues. NMR in Biomedicine 2006; 19: 765-780.
9. McAteer MA et al. In vivo magnetic resonance imaging of acute brain inflammation using microparticles of iron oxide. Nature Medicine 2007; 13: 1253-1257.
10. Blamire AM. The technology of MRI - the next 10 years? British Journal of Radiology 2008; 81: 601-617.
For a more in-depth discussion about future MRI developments, the reader is directed to the recent review by the same author .
Andrew M. Blamire, B.Sc., Ph.D.
Professor of MR Physics,
Newcastle upon Tyne,