Human whole-body MRI magnets for clinical use have room-temperature bores ~90 cm and magnetic fields up to 7 T. Research can be carried out in systems providing up to 11.7 T. These magnets typically use a NbTi superconducting (SC) wire. This material will not superconduct beyond ~12 T. 

Numerous SC magnets have been built and operated at fields >12 T, typically using Nb3Sn wire. The first SC magnet put into service at >12 T was in 1973 and 20 T was reached in 1990. The largest market for Nb3Sn magnets is nuclear magnetic resonance (NMR) for which 5 cm is a common size for the room-temperature bore. Nb3Sn-based NMR magnets up to 23.4 T have been available since 2010. Nb3Sn is also used in small bore magnets for condensed matter physics at high fields, pre-clinical MRI magnets, ion-cyclotron-resonance magnets, and a few other applications. 

In addition, Nb3Sn has been used for a few resistive-superconducting hybrid magnets for condensed matter physics that provide fields up to 45 T in a 3 cm room-temperature bore. But the SC parts of these magnets provide fields up to 14 T in a bore up to 70 cm (1999, MagLab). A few magnets have been developed to confine plasmas and enable nuclear fusion that were tested at fields up to ~10 T in bores of a few meters.

In 1986 the first of the High-Temperature Superconducting (HTS) materials was discovered. In 2007 a test coil using a “second-generation” of HTS materials reached 27 T. A 32 T SC magnet using HTS materials reached full field at the MagLab in 2017, ushering in a new age of ultra-high field magnets using these materials. There are now at least thirteen NMR magnets in operation at fields greater than can be reached by Nb3Sn. At least one system has been tested to 30.5 T. 

In recent years, >40 companies have been founded and have raised capital a total of >$9B to develop fusion power plants. Several of them are focused on magnetically-confined technologies requiring magnets with fields in the range of 20 T with bores > 90 cm.

The challenges associated with developing a 14 T magnet suitable for human MRI using Nb3Sn or HTS materials are presented.

Last year, ISEULT 11.7T MRI scanner delivered its first in vivo images. It showed the potential of the device but also some limitations, especially when it comes to Whole Brain (WB) imaging. In this presentation, the latest developments needed to access reliable WB imaging at this field will be presented with new in vivo results.

First images acquired on human subjects at 11.7T have shown that it was possible to increase significantly the spatial resolution, at a given time budget, with the help of the stronger SNR available at this B0. Nevertheless, those images also showed how much it was difficult to obtain homogeneous spin excitations on a large region of interest within the Specific Absorption Ratio limits. These first experiments also rose the high sensitivity of these high-resolution images to physiological and uncooperative movements of the volunteers. 

Before to move forward, with the help of collaborators from all around Europe and beyond, some developments have been done to introduce a new coil, new pulse design to fix the problem of the heterogeneous excitations. A new 3D EPI T2* weighted sequence has been introduced to achieve whole brain coverage at 300um isotropic in about 5min. Finally, we introduce new strategies to fight either physiological or uncompliant motions.

Using this tools, preliminary results obtained recently on a new series of volunteers for anatomical and functional applications will be disclosed for the first time.

To be updated...

Metabolism is central  to human health and disease, yet capturing its dynamics in vivo has long remained a challenge.  Hyperpolarized magnetic resonance imaging (HP-MRI) overcomes this by considerably enhancing the signal of 13C-labeled molecules (such as [1-13C]pyruvate), transforming MRI into a real-time, non-invasive tool for metabolic imaging1.

The unprecedented speed, scalability and simplicity of Parahydrogen induced polarization-based (PHIP) HP-MRI make it well-suited for integration into preclinical research and clinical workflows across many fields, including oncology, cardiology, and neurologic disease and disorders. NVision is bringing PHIP HP-MRI to the market with POLARIS, allowing users to generate doses at the push of a button and making hyperpolarization routine.  This brings hyperpolarized imaging agents to a cross section of research questions that had previously been inaccessible.

In oncology, hyperpolarized imaging agents provide in vivo insights for the validation of disease mechanisms, metabolic phenotyping, and real-time monitoring of treatment response2. In radiation therapy, they can serve as promising biomarkers throughout the clinical workflow, from evaluation of tumor sensitivity during planning to detecting early treatment response for potential treatment adaptation3. In neurology, these agents enable the detection of subtle metabolic changes in the brains of patients with traumatic brain injury4 or allow to monitor neuroinflammation5.  

NVision POLARIS is integral to the deployment of  hyperpolarized MRI, and reshaping our ability to visualize metabolism – as it happens.

References: 

1. Wang, Z. J. et al., Radiology 291 (2019). 

2. Chia, M.L. et al., Oncogene 44 (2025). 

3. Lee, C.Y. et al., J Neurooncol 152 (2021).

4. Hackett E.P. et al., iScience 23 (2020). 

5. Guglielmetti C. et al., Sci Rep 7 (2017).7).

To be updated...

Imaging at ultrahigh fields (UHF) provides increased resolution allowing for improvements in neuroscientific and clinical imaging. However, the associated decrease in RF wavelength at UHF causes wave interference within the imaged sample, resulting in transmit signal voids and specific absorption rate (SAR) hotspots. Currently, the best method to mitigate wave interference is parallel transmission (pTx), requiring coil arrays where each channel transmits the RF pulse with a different phase and magnitude. Although effective, pTx requires intricate coil arrays and resource-intensive calculations to determine the optimal phases and magnitudes for each patient. It is thus worth exploring a volume coil (as used for lower B0) because of its simplicity. We hypothesize that our MTS head coil can extend the application range of volume resonators to 10.5T with acceptable homogeneity and SAR performance.

This study investigates how skeletal maturation influences 23Na MRI relaxation parameters and the accuracy of tissue sodium concentration (TSC) quantification in normal human knee cartilage at 10.5T. The increased signal-to-noise ratio and prolonged long T2* relaxation component at higher field strengths significantly benefit 23Na MRI of ordered tissues such as cartilage. Twelve pediatric knee specimens (age range: 3 months to 11 years) were imaged with a whole-body 10.5T MRI system and a birdcage knee coil using a density-adapted 3D radial projection sequence. 23Na MRI parameters, including B1+, T1, biexponential T2*, and TSC, were calculated and compared with the reference biochemical assessments of cartilage composition from osteochondral biopsies. 23Na MRI on a 10.5 T system allowed accurate quantification of TSC in human knee cartilage throughout the process of cartilage maturation, as validated by the biochemical analysis of cartilage biopsies. The 10.5T system enabled accurate quantification of TSC throughout cartilage maturation, as validated by the biochemical analysis. Furthermore, 23Na relaxation parameters sensitive to 23Na dynamics provided useful insights into changes in collagen matrix composition during pediatric cartilage maturation. Significant changes in 23Na T2* relaxation parameters were observed between birth and five years of age, indicating that age-specific corrections for 23Na relaxation are necessary for accurate TSC quantification during early cartilage development. On the other hand, the relative stability of cartilage relaxation parameters after the age of 5 years supports the use of previously reported 23Na relaxation parameters from adult cartilage for accurate TSC quantification eliminating the need for time-consuming individual 23Na relaxation measurements in older children. These findings also suggest that compositional changes due to degenerative processes, such as osteoarthritis, may affect the accuracy of TSC quantification. While relevant for cartilage MRI at clinical field strengths, the present study also demonstrates the translational potential of 10.5T MRI. Future in vivo 23Na MRI studies at 10.5T may offer new insights into cartilage maturation, regeneration, and degenerative diseases such as osteoarthritis.

Brain–body interactions virtually regulate all aspects of life and behavior of vertebrate animals, including essential functions such as heart rate, breathing, digestion, and stress responses. Non-invasive monitoring of the dynamics of such complex interactions is however challenging, primarily due to the lack of adequate functional imaging modalities that can monitor the body and the brain simultaneously with sufficient spatial and temporal resolution. This challenge motivates the development of new imaging approaches that can capture activity not only of the brain but also of other organs, along with broader physiological processes. In this talk, I will present recent developments demonstrating that techniques alternative to the standard blood oxygenation level (BOLD) contrast such as ultrashort and zero echo time (UTE/zero-TE) MRI make it possible to expand functional MRI (fMRI) investigations “beyond the brain”. Leveraging their resilience to susceptibility artifacts, UTE/zero-TE techniques have indeed enabled for the first fMRI of the nose, a previously inaccessible region that is pivotal for studying peripheral–central nervous system interactions. Finally, I will demonstrate that UTE/zero-TE approaches offer unique advantages for investigating neurofluid dynamics such as cerebral blood flow and cerebrospinal fluid movement, without the complex oxygenation confound inherent to classical BOLD fMRI contrast. Together, these advances highlight the potential of UTE/zero-TE imaging to broaden functional imaging, offering new opportunities to study the nervous system in health and disease."

Comprehensive evaluation of the central nervous system (CNS), rather than isolated studies in the brain or spinal cord, are crucial for applications like pain, spinal cord injury, and neurodegenerative diseases. Simultaneous cortico-spinal imaging can disentangle circuits involved in different conditions by exploring interactions between spinal and brain regions. However, fMRI of the CNS faces challenges due to the requirement for magnetic field uniformity across a large field-of-view (FOV) or several FOVs. Simultaneous fMRI of the brain and distant spinal cord is currently unattainable, though studies have imaged the brain or brainstem with nearby cervical spinal cord regions. Proposed dynamic shimming approaches have drawbacks, including extended scanning time and limited coverage only to the cervical spinal cord. Sequential rather than simultaneous acquisitions of the brain and spinal cord also hinder comprehensive connectivity analyses of the entire CNS.

A radial zero-TE fMRI approach overcomes issues of magnetic field inhomogeneity and motion sensitivity seen in current echo planar imaging (EPI)-based methods. Zero-TE pulse sequences, with their lack of acquisition delay and phase encoding, and high bandwidths, naturally address these issues. The feasibility of a zero-TE approach using Multi-Band SWeep Imaging with Fourier Transformation (MB-SWIFT) has been demonstrated for detecting brain activation in human studies and has shown numerous benefits for studying rat brain function. Zero-TE fMRI contrast originates from the inflow of unsaturated blood differentiating it from traditional blood oxygen level dependent (BOLD) contrast. We have demonstrated this technique to work robustly in spinal cord fMRI studies with spinal cord stimulation in rats. Recently, we also showed that this sequence can acquire fMRI signals from the brain and lumbar spinal cord simultaneously without dedicated shimming strategies.

Functional magnetic resonance imaging (fMRI) relies on the blood oxygen level–dependent (BOLD) signal, which reflects changes in local cerebral blood flow. Yet, the cellular and molecular mechanisms that generate and sustain functional hyperemia remain incompletely understood. Using two-photon imaging in awake mice, we identify a previously unrecognized bimodal structure of sensory-evoked hyperemia: an early transient dilation of arterioles, followed by a sustained amplification when neuronal activity persists. This late component critically depends on astrocytic Ca²⁺ signaling. Clamping astrocyte Ca²⁺ with a plasma membrane Ca²⁺ ATPase (CalEx) selectively reduces sustained, but not brief, dilation. Conversely, chemogenetic elevation of astrocytic Ca²⁺ augments sustained hyperemia. Pharmacological inhibition of NMDA receptors or blockade of epoxyeicosatrienoic acid synthesis also diminishes the late, sustained phase, without altering the early response. These results establish a specific role for astrocytes in amplifying blood flow during prolonged neuronal activation, providing a mechanistic substrate for the sustained component of the BOLD signal.

To be updated...

Field monitoring, the direct measurement of the dynamic magnetic field, has been in use for over 15 years. During this time there has been significant work demonstrating its utility and potential for a diverse range of applications from improved image reconstruction for diffusion and fMRI, characterizing new MR scanners and gradients, and accelerating pulse sequence development. Recent examples of field monitoring success stories will be shared, showcasing how the Skope community has harnessed field monitoring for neuroscience, body imaging, and methods development.

As applications of field monitoring evolve, so to have the hardware platforms. Two new hardware platforms, NYOX and the NC7T, represent developments which address the needs of the UHF, Neuroscience, and MR methods development communities. NYOX is our brand-new platform for field monitoring featuring fully redesigned hardware and software architectures. These new architectures enable streamlined workflows, improved system performance, and continues the Skope tradition of open raw data formats. The NC7T is our 7T RF coil designed with fully integrated field monitoring capabilities. Designed in collaboration with MRCoilTech, this coil enables high acceleration and SNR via its 64 receive channels. The fully integrated field monitoring capabilities enables users to field monitor experiments without additional hardware and without changes to their subject handling workflow.

The push toward ultra-high field MRI has created demand for robust, high-channel radiofrequency coils that can deliver both high signal-to-noise ratio and reliable performance under the stresses of peak gradient and RF conditions. These challenges are particularly acute in regions such as the cerebellum, where conventional head coils may struggle to provide adequate coverage and sensitivity. I will present recent progress on a head coil designed to overcome these limitations. Parallel versions are being developed at UWO and UTSW for Siemens and Philips platforms, providing a consistent foundation for multi-site cerebellar imaging. Guided by ultimate intrinsic SNR simulations, the coil features a 32-channel receive array tailored to the posterior head and neck, with a close-fitting design and large anterior elements that boost sensitivity in the cerebellum and brainstem while preserving comfort and access for behavioral tasks. A two-row dipole/loop transmit configuration was chosen to improve parallel transmission performance in the cerebellum. Gradient-induced eddy currents and RF coupling are addressed through careful housing design and redundant detuning strategies, ensuring safe operation at high duty cycles and with spatially selective pulses. This work demonstrates strategies for developing high-channel coils that withstand the mechanical and electromagnetic demands of daily use, while expanding the possibilities for high-quality cerebellar imaging at 7T.

To be updated...

To be updated...

Image quality and resolution in MRI are fundamentally constrained by the performance of radiofrequency (RF) coils used to excite spins and receive signal. Electromagnetic (EM) simulations are essential for predicting and optimizing coil performance; however, existing tools are often slow, memory-intensive, not tailored to MRI, and require expensive licenses. My talk will introduce an open-source EM simulation toolbox purpose-built for RF coil design in MRI that addresses the above-mentioned limitations. The toolbox consists of four complementary software components: (1) a full-wave 3D EM solver based on the wire-surface-volume integral equation, accelerated via tensor decomposition to reduce memory usage of simulations involving fine body model resolutions; (2) a reduced-order model technique enabling evaluations of multiple coil designs for the same body model in minutes; (3) a fully automated circuit co-simulator for coil tuning, matching, decoupling, preamplifier decoupling, and detuning; and (4) an EM basis generator for computing ultimate intrinsic and optimal performance metrics. The educed-order model technique computed the coil currents for a 31-channel 7T head coil at least 70 times faster relative to a commercial solver. The co-simulator automatically tuned the coil in less than 40 minutes, comparing to a few days time needed in the commercial package. The average difference in signal-to-noise ratio (SNR) performance was less than 15% over the entire head. We demonstrated the toolbox with a series of simulations for different applications at 7T: A shielded brain transceiver, three prostate transceive and receive arrays loading a full body model, and a cervical spine and cerebellum array with separate transmit and receive components. The proposed open-source toolbox enables fast, memory friendly, accurate, and anatomy-specific RF coil array design, making it a powerful tool for advancing RF coil engineering.