Jerome Allard1, Alexis Amadon2, Guy Aubert1, Jean Belorgey1, Christophe Berriaud1, Cédric Bonnelye2, Romain Boudoin1, Nicolas Boulant2, Thierry Boussuge1, Fawzi Boumezbeur2, Philippe Brédy1, Peter Dietz3, Guillaume Dilasser1, Yannick Drouen1, Olivier Dubois1, Sandrine Faict-Bastin1, Ignacio Gonzalez Insua3, Vincent Gras2, Quentin Guihard1, Patrice Guiho1, Jean-Christophe Guillard1, Vincent Jannot1, François-Paul Juster1, Felix Koeber3, Brice Koestel4, Denis Le Bihan2, René Leboeuf1, Cécile Lerman2, Frédéric Leprêtre2, Ange Lotodé1, François Nunio1, Lionel Quettier1, Hermann Landes5, Hervé Lannou1, Aurélien Massire4, Franck Mauconduit2, Frédéric Molinié1, Hubert Neyrial1, Cédric Péron1, Luc Renou4, Arnaud Roger1, Philippe Rouffiat4, Wolfram Ruth3, Thierry Schild1, Patrick Sieber3, Mathieu Santin6,7, Martin Schroeder3, Loris Scola1, Armand Sinanna1, Vadim Stepanov1, Mathieu Szmigiel1, Olivier Tellier1, Robert Touzery1, Pierre Védrine1, Alexandre Vignaud2, Christian Walter1, and Karsten Wiclow3
1University of Paris-Saclay, CEA, Irfu, Gif sur Yvette, France, 2University of Paris-Saclay, CEA, CNRS, BAOBAB, NeuroSpin, Gif sur Yvette, France, 3Siemens Healthcare GmbH, Erlangen, Germany, 4Siemens Healthcare SAS, Saint-Denis, France, 5Ensimtech, Buckenhof, Germany, 6Sorbonne University, Institut du Cerveau - Paris Brain Institute - ICM, INSERM, CNRS, Paris, France, 7ICM, Centre de NeuroImagerie de Recherche – CENIR, Paris, France
Synopsis
The Iseult project
started in 2001 and is a collaborative effort involving CEA and the University
of Freiburg in academics, Guerbet, Siemens Healthineers and Bruker Biospin as
industrial partners. One central aspect of the project has been the design by
CEA of a whole-body magnet of 11.7T field strength with a 90-cm wide bore.
After nearly 20 years of research and development, prototyping, integration and
tests, first images have been acquired with this unique MRI scanner. Some key
validation results are presented here.
Introduction
MRI at ultra-high field (UHF) offers great
promises in SNR and CNR and has fueled the race to achieve ever higher fields,
the highest field used on humans to date being 10.5T at the University of
Minnesota (CMRR). With higher fields however come greater challenges (e.g. field
homogeneity and stability) which must be overcome to enable proper exploitation.
Here we review the key results of the tests performed in the last year to
finalize the commissioning of this unique 11.7T whole-body MRI scanner1,
and we present its first high resolution images acquired on a pumpkin and on an
ex-vivo brain.Methods
The main superconducting coil of the Iseult
magnet is made of NbTi and is immerged into a 7000-liter He vessel at a
temperature of 1.8 K (below the 2.17 K transition from liquid to superfluid He).
Following a double pancake design, it sustains a current of 1483 A and has an
inductance of 308 H, allowing to ramp-up the field from 0 to 11.7T in
approximately 5 hours. This magnet is actively shielded by two 5-m diameter
superconducting coils also immerged in superfluid He. The whole magnet assembly
weighs 132 tons with external diameter and length of 5 m. After the cool down
of the magnet and the first ramp-up at 11.7T in 2019, the imaging equipment (of
the MAGNETOM 11.7T MRI series) has been integrated around the magnet. Gradient
coil-magnet interaction tests were performed at 3T, 7T, 10.2T and 11.7T, the
gradient being the Siemens Healthineers SC72 whole-body gradient. They included acoustics,
vibration and power deposition measurements during variable speed frequency
sweeps of the three gradient coils independently. Figure 1 reports some
vibration data measured on the gradient, the lead tube (designed by CEA to
screen gradient-magnet interactions at specific frequencies2) and
the magnet cryostat. Temporal field stability (spec of 0.05 ppm/hour max at
rest) was measured at 11.7T with an FID sequence over 14 hours on a phantom.
The External Interference Shield (EIS) designed to screen external field
perturbations was tested with a Skope (Skope MRT, Zurich, Switzerland) field
camera with measurements performed at an intermediate field of 7T every 105 ms
while a bus was driving on the road nearby. Field homogeneity (spec of 0.5 ppm
peak-to-peak over the surface of a 22 cm diameter sphere with passive and cryo
shimming only) was mapped by a Metrolab (Metrolab, Geneva, Switzerland) field camera.
Tune-up and Quality Assessment of the imaging equipment, in quadrature mode
only for the moment, was performed by Siemens Healthineers with support of CEA.
First images were acquired on a pumpkin and an ex-vivo brain using a Gradient
Echo sequence with a 0.4 mm isotropic spatial resolution (TR=20 ms, TE = 1.82/2.5
ms, Matrix = 512 × 512 × 448, 4 NEX, TA = 5 H 05 min), using the single channel
service quadrature coil (QED) delivered by Siemens Healthineers. Results
Field stability measurement
is shown in Fig. 2a revealing a fluctuation of less than 0.01 ppm over 14 hours.
The EIS also reduces by 10 fold the field disturbance caused by a bus driving
on the road 50 m from the magnet (Fig. 2b). Field homogeneity relies on passive
shim iron (optimized schematics of Al/Iron blocks), cryoshims and the Siemens Healthineers active shim.
The last measurement returns a peak-to-peak variation of 0.88 ppm and a RMS
deviation of 0.15 ppm with the passive and (partial) cryo shim only (Fig. 3), the
nominal frequency being 499.415 MHz. Despite a severe but expected3
RF field inhomogeneity at 500 MHz with a quadrature coil, these experimental
conditions led to the visualization of exquisite details on high spatial
resolution images of the pumpkin (e.g. vegetal fibers) and of the ex-vivo brain
(e.g. internal structure of the thalamus), thus validating up to this stage the
commissioning of the 11.7T Clinical MRI system. Discussion and conclusion
The results reported
in this work reflect and validate the research and developments performed at
CEA for nearly 20 years and now make MRI at 11.7T and with a 90 cm diameter
bore a reality. The safety assessment of this new medical device will be
started soon to seek authorization to scan human volunteers. Parallel
transmission also will be deployed in the near future to mitigate the RF field
inhomogeneity problem. Small refinements on the passive shim configuration are also needed
to reach the final specifications.Acknowledgements
The Musée National
d’Histoire Naturelle (Marc Herbin) in Paris is thanked for lending CEA the
post-mortem brain for scanning. This part of the Iseult project has been funded
by CEA and BPIfrance. NB is funded by the European Union’s Horizon 2020 research
and innovation program under grant agreement No 885876 (AROMA project). References
[1] D. Le Bihan, T.
Schild, Human Brain MRI at 500MHz, scientific perspectives and technological
challenges, Supercond. Sci. Techno. 2017;30:1-19.
[2] G. Aubert. NMR
imaging system with reduced cryogenic losses and reduced acoustic noise. US patent
no 8,410,777 B2.
[3] A. Webb. Concepts
in Magnetic Resonance Part A, 2011; 38A:148-184.