C.D.E. Van Speybroeck1, W. Roskamp1,2,3, R.S. Vinke2, M. van der Graaf1, and W.M. Brink3
1Department of Imaging, Radboud University Medical Center, Nijmegen, Netherlands, 2Donders Institute for Brain, Cognition and Behaviour, Department of Neurosurgery, Radboud University Medical Center, Nijmegen, Netherlands, 3Magnetic Detection and Imaging group, TechMed Centre, University of Twente, Enschede, Netherlands
Synopsis
Keywords: Safety, Safety, Deep Brain Stimulation; Parallel Transmission
Motivation: Bilateral deep brain stimulation (DBS) surgery would benefit from safe intraoperative 3T MRI using conventional scanners.
Goal(s): Our goal was to develop a method to determine patient-specific safety constraints to scan a bilateral lead configuration with minimal RF-heating, by minimizing RF-induced currents at maximal B1.
Approach: We scanned a phantom with two leads in different configurations, measured local heating and determined the transimpedance.
Results: We could determine the phantom-specific safety constraints and find a low-SAR sequence meeting the constraints, after optimizing the two channel RF-shimming to minimize RF-heating by minimizing RF-induced currents at maximal B1.
Impact: The proposed method can be used to minimize
heating of a bilateral lead configuration and estimate patient-specific safety
constraints, but further research has to be conducted to be able to use it
during deep brain stimulation surgery.
Purpose
Deep brain
stimulation (DBS) is a well-established surgical treatment to relieve symptoms
of many neurological disorders, including Parkinson’s disease. Accurate
positioning of the leads in the subthalamic nucleus (STN) renders 3T MRI the
preferred clinical modality for providing intraoperative feedback. Currently, most DBS systems are not yet labeled at 3T due to the
increased risk of RF-heating, and this holds in particular for partially
implanted DBS systems in the intraoperative setting. This becomes
increasingly limiting on 3T protocols in DBS patients due to increasing regulatory requirements and associated costs of labelling.
A potential solution to manage RF-safety in DBS
configurations is parallel transmission, where two-channel RF-shimming has been
shown to be able to mitigate the RF-induced current on single-lead
configurations.1,2 Clinical DBS implantation typically involves
bilateral lead placement, however, for which the two degrees of freedom
(relative RF amplitude and phase) are known to be insufficient for complete
nulling of both currents simultaneously.3,4 In this study, we explore
the potential of minimizing the RF-induced currents on both leads with
two-channel RF-shimming, while maximizing the available B1+,
and assess the potential of performing a low-SAR protocol based on the obtained
patient-specific safety constraints.Methods
A phantom (30g/L
agar powder; 2g/L NaCl; 16×13×12cm), see figure 1A, mimicking the brain, including electrical
conductivity and thermodynamics, containing two parallel, centrally placed
directional DBS leads (Vercise Cartesia Directional, DB-2202, Boston
Scientific, Marlborough, MA) was created. To record the heating, three fiber-optic
temperature probes (OTP-M series, Opsens, Quebec City, QC) were used (one at
each tip, one as reference).
The phantom was
scanned following the workflow described in figure 1B. First, low-SAR channelwise
GRE and B1+-data is acquired similarly to Eryaman,
et al.2 The DBS lead position and Tx-Null artefact are characterized
by an outlier detection procedure, yielding quantitative values for the
channelwise induced current, as shown in figure 2A. The combined implant
friendly (IF) mode was then determined by maximizing the B1+
divided by the maximum lead current I, as shown in figure 2B. An
additional low-SAR GRE and B1+-dataset is acquired
using these excitation settings for confirmation.
Local heating
was measured using fiber-optic temperature sensors directly adjacent to the
lead electrodes in each of their individual IF mode, as well as the combined (IF)
mode and quadrature mode (time = 3:04,
FA = 110°, TR = 2600ms, TE = 110ms, slices = 26, slice thickness = 2mm). Following the method of Sadeghi-Tarakameh, et al.5, the remaining induced current I was translated to
a patient-specific safety constraints, limiting the temperature increase to 2°C, using the transimpedance RDBS.6Results
The 2D GRE calibration
scans (Cal 1-3), shown in figure 3A, were used to determine the top, bottom and
combined IF mode. As shown in figure 3B, the Tx-null artifacts of both leads have
a similar magnitude in the IF combi mode, indicating equality of the induced
current.
Figure
4 shows the heating curves, induced by the TSE sequence, for the top lead (A)
and the bottom lead (B) of the IF (top), IF (bottom), IF (combi) and quadrature mode, as a reference, indicating that the IF combi mode results in
minimal heating in both leads. This configuration limits the temperature rise to
0.4°C in both leads.
From the obtained phantom-specific transimpedance RDBS and the remaining induced currents I for both leads in IF combi mode,
the B1+rms-limit of the top lead was determined to be 0.81µT and of
the bottom lead 0.65µT, complying with the B1+rms = 0.60µT of
the low-SAR protocol, in contrary to the Standard of Care (SoC) protocol with a
B1+rms of 2.40µT. Images aquired with both protocols are shown in figure 5.Discussion
To translate the
phantom study to patients, reliability of the patient-specific safety constraints
should be increased by including error margins and scanning more orientations. Also
further studies should be conducted, starting with a patient-like phantom with
the leads implanted in a patient-like manner (angulated and with the tips
closer together), and concluding with a cadaver study, to provide the most realistic
setting with correct electrical conductivity, thermodynamics and permittivity.
In the end, the
created low-SAR protocol will not be one-size fits all: the scan time can be
reduced by increasing the TR without influencing the T2-contrast. Conclusion
In this study,
we have explored the potential of minimizing the RF-induced currents on a
bilateral DBS configuration using two-channel RF-shimming. We showed that the combined
IF mode produces the least heating in both leads, were able to determine the phantom-specific
B1+rms-limit and create a complying low-SAR protocol showing comparable
image quality to the SoC protocol.Acknowledgements
We would like to
thank Kristian Overduin, Andor Veltien and Dagmar Grob from the Radboudumc for
their contribution to the project, and Yigitcan Eryaman of the CMRR for helpful discussions. References
- Eryaman Y, Turk EA, Oto C, et
al. Reduction of the radiofrequency heating of metallic devices using a
dual-drive birdcage coil. Magn Reason Med. 2013;69(3):845-852.
- Eryaman Y, Kobayashi N, Moen S,
et al. A simple geometric analysis method for measuring and mitigating RF
induced currents on Deep Brain Stimulation leads by multichannel
transmission/reception. NeuroImage. 2019;184:658‐668.
- Etezadi-Amoli
M, Stang P, Kerr A, et al. Controlling radiofrequency-induced currents in
guidewires using parallel transmit. Magn Reason Med. 2015;74(6):1790-1802.
- Sadeghi-Tarakameh A, DelaBarre
L, Zulkarnain NIH, et al. Implant-friendly MRI of deep brain stimulation
electrodes at 7 T. Magn Reason Med. 2023;90:2627-2642.
- Sadeghi-Tarakameh A, Zulkarnain
NIH, He X, et al. A workflow for predicting temperature increase at the
electrical contacts of deep brain stimulation electrodes undergoing MRI. Magn
Reason Med. 2022;88:2311-2325.
- Finelli DA, Rezai AR, Ruggieri
PM, et al. MR Imaging-Related Heating of Deep Brain Stimulation Electrodes: In
Vitro Study. Am J Neuroradiol. 2002;23(10):1795-1802.