Sebastian Gantz1,2, Volker Hietschold3, Sergej Schneider1,2,4, and Aswin Louis Hoffmann1,2,5
1Institute of Radiooncology-OncoRay, Helmholtz-Zentrum Dresden - Rossendorf, Dresden, Germany, 2OncoRay – National Center for Radiation Research in Oncology, Faculty of Medicine and University Hospital Carl Gustav Carus, Technische Universität Dresden, Helmholtz-Zentrum Dresden-Rossendorf, Dresden, Germany, 3Department of Radiology, Faculty of Medicine and University Hospital Carl Gustav Carus, Technische Universität Dresden, Dresden, Germany, 4Technische Universität Dresden, Carl Gustav Carus Faculty of Medicine, Dresden, Germany, 5Department of Radiotherapy and Radiation Oncology, Faculty of Medicine and University Hospital Carl Gustav Carus, Technische Universität Dresden, Dresden, Germany
Synopsis
The targeting precision of proton therapy is expected to benefit from
real-time MRI guidance. We developed a setup of a first prototype in-beam MRI scanner with a proton pencil
beam scanning nozzle. Dipole magnets in the nozzle used for beam steering
produce time-dependent magnetic fringe fields that may interfere with the MR
image acquisition. In this study, we show that vertical beam steering shows no
degradation of the MR image quality, whereas horizontal beam steering
introduces severe ghosting artefacts in phase encoding direction. The origin of
these artefacts is unraveled and strategies to eliminate or correct these
artefacts are proposed.
Introduction
Real-time
MRI guidance is expected to significantly improve the targeting precision of proton
therapy (PT) 1. Hybrid systems integrating MRI and PT do not exist
due to a number of hitherto open technological questions. To assess the mutual
magnetic interactions between these systems, we recently realized a setup of a
first prototype in-beam MRI scanner
at a horizontal proton pencil beam scanning (PBS) research beamline. The static
and dynamic magnetic fringe fields produced by the PBS beamline may interfere
with the imaging fields of the MRI scanner. The aim of this study was to
investigate the effects of the dynamic fringe fields of the PBS beamline on the
MR image quality during simultaneous proton beam irradiation and MR image
acquisition. Methods
A 0.22 T C-shaped open MR scanner (MrJ2200, ASG Superconductors S.p.A.,
Genua, Italy) was combined with a horizontal PBS research beamline connected to
an isochronous cyclotron (C230, Ion Beam Applications SA, Louvain-la-Neuve,
Belgium). The scanner was mounted on a mobile transport platform and enclosed
by a Faraday cage made of plywood panels and a 120 µm copper foil (Fig. 1) 2.
Volumetric irradiation is achieved by magnetic beam steering in horizontal and
vertical direction by a pair of dipole magnets (scanning magnets), whereas the
depth variation is determined by the initial energy of the proton beam.
To investigate the time-dependent beam steering effects on the MR image
quality a proton beam of 220 MeV was subsequently scanned along either a horizontal
or vertical line of 40 cm and 20 cm length, respectively, using 9, 17, 33 and
81 equidistant dose spots. The total irradiation time of 20 s was matched to
the acquisition time of a single-slice gradient echo (GRE) sequence (TE=20 ms,
TR=80 ms, FA=60°, FOV=20×20 cm2, acquisition matrix=256×256, bandwidth=42 Hz/px,
Cartesian sampling) that was used to image a homogeneous transversal slice of
the ACR Small Phantom 3, being positioned inside a knee receiver
coil in the center of the field-of-view of the MR scanner.
Imaging and irradiation was manually synchronized such that image
acquisition started ~1 s before beam scanning, to assure that the pre-image
frequency calibration was unaffected by magnetic field changes due to the scanning
of the proton beam. The MR image quality was qualitatively evaluated and
compared to reference images acquired without beam scanning but with the beamline
energized.Results
MR images acquired during vertical beam scanning showed no visual differences
to reference images (Fig. 2a-b). This indicates that vertical beam scanning does
not deteriorate the MR image quality. The MR images acquired during horizontal
beam scanning, however, are blurred and show coherent ghosting artefacts in the
phase encoding (PE) direction (Fig. 2c-f). The number of ghosts in the images is
inversely proportional to the number of dose spots scanned. This indicates a
phase mismatch between adjacent lines in k-space,
caused by a change of the central resonance frequency due to the fringe field
of the beam scanning magnets. This phase shift is recognized in the phase map
of the k-space data (Fig. 3) and
coincides with the position change between dose spots.Discussion
For the first time, the effects of the dynamic fringe fields of a proton
PBS beamline on the MR image quality have been assessed during simultaneous
proton beam irradiation and MR image acquisition. Severe ghosting artefacts
were observed in a GRE image acquired during proton beam scanning in the
horizontal direction only. The ghosting artefacts exhibit a systematic behavior
that is related to the time structure of the dose spots delivered. The phase
maps of the k-space data prove that the
artefacts are caused by phase offsets between adjacent lines, which can be
understood by phase accumulation due to spot positon-dependent changes in the resonance
frequency. For each change in spot position the static magnetic field of the MR
scanner is modulated by the overlapping magnetic fringe field of the nearby
beam scanning magnets. Since the fringe field of the horizontal scanning magnet
is parallel to the main field of the MR scanner, its impact on the central
resonance frequency is larger than that of the vertical scanning magnet, whose
fringe field is orthogonal to the main field of the MR scanner.
To eliminate the ghosting artefacts, either
magnetic decoupling/shielding or online/offline correction strategies for image
reconstruction have to be deployed. Online corrections can possibly be achieved
through a navigator-echo approach assessing the central frequency in each k-space line 4. For offline corrections,
either a phase-error correction strategy 5 or a deep-learning based
image correction approach 6 could be used. Ongoing investigations
must show which of these approaches turns out to be the most effective. Conclusion
Simultaneous image acquisition and proton beam irradiation with an in-beam MR scanner at a proton pencil
beam scanning beamline produces severe MR image ghosting artefacts when the
beam is steered in a direction perpendicular to that of the main imaging field.
The origin and nature of the artefacts is well understood and appropriate means
for elimination of the artefacts have to be implemented and studied in future
investigations.Acknowledgements
No acknowledgement found.References
1.
Oborn BM,
Dowdell S, Metcalfe PE, et al. Future of medical physics: Real‐time MRI‐guided
proton therapy. Med. Phys. 2017; 44(8): e77-e90.
2.
Schellhammer SM, Hoffmann AL, Gantz
S, et al. Integrating a low-field open
MR scanner with a static proton research beam line: proof of concept. Phys. Med. Biol. 2018; 63(23):
23LT01.
3.
American College
of Radiology. Phantom Test Guidance for Use of the Small MRI Phantom for the
MRI Accreditation Program. April 17, 2018
https://www.acraccreditation.org/-/media/ACRAccreditation/Documents/MRI/SmallPhantomGuidance.pdf?la=en. Accessed October 10, 2019.
4.
Li J, Wang Y, Jiang Y, et al. Image correction during large and rapid B0
variations in an open MRI system with permanent magnets using navigator echoes
and phase compensation. Magn. Reson. Imaging 2009; 27(7): 988-993.
5.
Broche LM,
Ross PJ, Davies GR, et al. Simple algorithm for the correction of MRI image
artefacts due to random phase fluctuations. Magn. Reson. Imaging 2017; 44:
55-59.
6.
Küstner T, Armanious K, Yang J, et
al. Retrospective correction of
motion‐affected MR images using deep learning frameworks. Magn. Reson. Med.
2019; 82(4): 1527-1540.