In-silico analysis of motion-informed dose adjustment strategies for TMS-fMRI
Poster No:
120
Submission Type:
Abstract Submission
Authors:
Sarah Grosshagauer1, Michael Woletz1, Maria Vasileiadi1, Martin Tik2, Christian Windischberger3
Institutions:
1Medical University of Vienna, Vienna, Vienna, 2High Field MR Center, Center for Medical Physics and Biomedical Engineering, Medical University of V, Vienna, Austria, 3Medical University of Vienna, Vienna, Austria
First Author:
Co-Author(s):
Martin Tik
High Field MR Center, Center for Medical Physics and Biomedical Engineering, Medical University of V
Vienna, Austria
High Field MR Center, Center for Medical Physics and Biomedical Engineering, Medical University of V
Vienna, Austria
Introduction:
Concurrent TMS/fMRI leverages the high temporal resolution of TMS interventions and the spatial resolution of BOLD-fMRI, yet, motion remains a major challenge. Using real-time motion monitoring, we could previously detect that participants moved significantly throughout a concurrent TMS/fMRI session, resulting not only in reduced image quality, but also in changes of the induced electric field (E-field) in the targeted brain region. Thus, we previously developed a framework for real-time adjustment of stimulation intensity based on motion tracking (figure 1a). Within this study we evaluate methodologies for dose-adjustments by performing in-silico comparisons between distance-correction, vector-potential informed corrections and E-field simulation.
Methods:
We used SimNIBS 4.0.0 (Thielscher et al. 2015) and the MagVenture MRi-B91 coil model to simulate many different coil orientations and positions. Simulations were targeted to the hand knob on M1 (MNI: -40 -20 42 (Cárdenas-Morales et al. 2014)), mapped to the space of the individual head model. Samples were defined within a cylindrical volume of interest (VOI, r=10 mm, h=10 mm) to simulate the potential range of motion during TMS/fMRI. In addition to coil position, we included coil orientation (rotation/tilt) by applying a uniform deterministic sampling of the 3D-rotation group SO(3) using Hopf fibration (Yershova et al. 2010). All obtained poses were checked for plausibility, e.g. that the TMS coil did not intersect with the headmesh. Using the fast auxiliary dipole technique (Gomez et al. 2021), we obtained the E-field magnitude within the target ROI for all plausible poses. The optimal coil pose and corresponding E-field were defined as reference (figure 1b).
Subsequently, we performed an in-silico comparison between different dose adjustment methods: distance-based correction for Euclidean and orthogonal distance between coil and target as proposed by (Stokes et al. 2005; 2007) (change in stimulation intensity of 2.9%/mm) and dose adjustment based on the vector potential of the TMS coil in the target ROI, which was calculated by transforming and interpolating the SimNIBS coilfile (Drakaki et al. 2022) according to the evaluated coil pose. We obtained adjustment factors for the stimulator output (maintaining the reference dose) for all methods and compared them to the change in E-field.
Subsequently, we performed an in-silico comparison between different dose adjustment methods: distance-based correction for Euclidean and orthogonal distance between coil and target as proposed by (Stokes et al. 2005; 2007) (change in stimulation intensity of 2.9%/mm) and dose adjustment based on the vector potential of the TMS coil in the target ROI, which was calculated by transforming and interpolating the SimNIBS coilfile (Drakaki et al. 2022) according to the evaluated coil pose. We obtained adjustment factors for the stimulator output (maintaining the reference dose) for all methods and compared them to the change in E-field.
·Figure 1
Results:
Kernel-density estimate plots of the calculated dose adjustments compared to E-field associated changes can be found in figure 2a. All methods underestimated correction factors compared to simulated E-field changes for most samples (figure 2b). Mean squared errors (MSE) compared to E-field guided adjustments were 0.12 (Euclidean) and 0.13 (z-distance) for distance based corrections and 0.09 for corrections based on vector potential. A subsample of poses with identical orientation compared to the reference position, i.e. pure translation of the coil, is plotted in figure 2c. In this case, MSE was 0.001 for distance based corrections but 0.003 if vector potential was used.
·Figure 2
Conclusions:
We successfully extended available E-field simulation software to allow for simulations in an extended volume including a homogeneous sampling of tilt and orientation of the coil. Comparisons of different dose adjustment methodologies revealed closest agreement with E-field based simulations if adjustment is based on interpolations of the vector potential, if translation as well as rotation/tilt of the coil is considered. For constant orientation, the distance based correction performed best. However, none of the available correction methods could capture the full extent of E-field changes as the tissue-interactions are simply not included. While simplified corrections might be valid if coil motion is small, a-priori E-field simulations and corrections based on these values are of utmost importance to obtain high consistency in target dose.
Brain Stimulation:
Non-invasive Magnetic/TMS
TMS 1
Modeling and Analysis Methods:
Methods Development 2
Motion Correction and Preprocessing
Novel Imaging Acquisition Methods:
BOLD fMRI
Keywords:
Data analysis
FUNCTIONAL MRI
Modeling
MRI
Transcranial Magnetic Stimulation (TMS)
Other - motion tracking, dose adjustment
1|2Indicates the priority used for review
Provide references using author date format
Drakaki, Maria (2022). “Database of 25 Validated Coil Models for Electric Field Simulations for TMS.” Brain Stimulation 15 (3): 697–706. https://doi.org/10.1016/j.brs.2022.04.017.
Gomez, Luis J., Moritz Dannhauer, and Angel V. Peterchev. 2021. “Fast Computational Optimization of TMS Coil Placement for Individualized Electric Field Targeting.” NeuroImage 228 (March): 117696. https://doi.org/10.1016/j.neuroimage.2020.117696.
Stokes, Mark G. (2007). “Distance-Adjusted Motor Threshold for Transcranial Magnetic Stimulation.” Clinical Neurophysiology 118 (7): 1617–25. https://doi.org/10.1016/j.clinph.2007.04.004.
Stokes, Mark G. (2005). “Simple Metric for Scaling Motor Threshold Based on Scalp-Cortex Distance: Application to Studies Using Transcranial Magnetic Stimulation.” Journal of Neurophysiology 94 (6): 4520–27. https://doi.org/10.1152/jn.00067.2005.
Thielscher, Axel (2015). “Field Modeling for Transcranial Magnetic Stimulation: A Useful Tool to Understand the Physiological Effects of TMS?” In 2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 222–25. https://doi.org/10.1109/EMBC.2015.7318340.
Yershova, Anna (2010). “Generating Uniform Incremental Grids on SO(3) Using the Hopf Fibration.” The International Journal of Robotics Research 29 (7): 801–12. https://doi.org/10.1177/0278364909352700.