Computational Artifacts of the In Situ Electric Field in Anatomical Models Exposed to Low-Frequency Magnetic Field

Gómez-Tames, José; Laakso, Ilkka; Haba, Yuto; Hirata, Akimasa; Poljak, Dragan; Yamazaki, Kazuhiko

doi:10.1109/temc.2017.2748219

Cited by 52 publications

(42 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…These metrics are applied to the different deep brain structures separately and also to the brain (white matter and grey matter) and the whole deep brain tissues. To mitigate numerical artifacts derived from computing the EF using the voxel model at the surface of the CSF-brain boundaries (Reilly and Hirata, 2016), post-processing based on the 99.9 th percentile value of the EF was applied for each tissue (Gomez-Tames et al, 2018). In this work, the EF strength was adopted as metric of neuromodulation to demonstrate the accuracy of the proposed segmentation.…”

Section: Analysis Methodsmentioning

confidence: 99%

End-to-end semantic segmentation of personalized deep brain structures for non-invasive brain stimulation

Rashed¹,

Gómez-Tames²,

Hirata³

2020

Neural Networks

Self Cite

View full text Add to dashboard Cite

Electro-stimulation or modulation of deep brain regions is commonly used in clinical procedures for the treatment of several nervous system disorders. In particular, transcranial direct current stimulation (tDCS) is widely used as an affordable clinical application that is applied through electrodes attached to the scalp. However, it is difficult to determine the amount and distribution of the electric field (EF) in the different brain regions due to anatomical complexity and high inter-subject variability. Personalized tDCS is an emerging clinical procedure that is used to tolerate electrode montage for accurate targeting. This procedure is guided by computational head models generated from anatomical images such as MRI. Distribution of the EF in segmented head models can be calculated through simulation studies. Therefore, fast, accurate, and feasible segmentation of different brain structures would lead to a better adjustment for customized tDCS studies.In this study, a single-encoder multi-decoders convolutional neural network is proposed for deep brain segmentation. The proposed architecture is trained to segment seven deep brain structures using T1-weighted MRI. Network generated models are compared with a reference model constructed using a semi-automatic method, and it presents a high matching especially in Thalamus (Dice Coeffi-cient (DC) = 94.70%), Caudate (DC = 91.98%) and Putamen (DC = 90.31%) structures. Electric field distribution during tDCS in generated and reference models matched well each other, suggesting its potential usefulness in clinical practice.

show abstract

Section: Analysis Methodsmentioning

confidence: 99%

End-to-end semantic segmentation of personalized deep brain structures for non-invasive brain stimulation

Rashed¹,

Gómez-Tames²,

Hirata³

2020

Neural Networks

Self Cite

View full text Add to dashboard Cite

show abstract

“…Numerical simulation is the most popular noninvasive approach to characterize the induced electric field (E‐field) in a human head during transcranial magnetic stimulation (TMS) [Deng et al, ]. Various computational methods have been proposed to solve equations for the E‐field in anatomical head models [Gomez‐Tames et al, ]. After discretization, solving the resulting system of linear equations is time‐consuming, usually consisting of 95% of the total time of the numerical analysis [Chen et al, ].…”

Section: Comparison For Time Cost Of Ilu‐bicgstab Using B‐febs and Thmentioning

confidence: 99%

An optimized block forward‐elimination and backward‐substitution algorithm for GPU accelerated ILU preconditioner in evaluating the induced electric field during transcranial magnetic stimulation

Wei

et al. 2019

Bioelectromagnetics

View full text Add to dashboard Cite

“…We defined the voxel SAR n as the n th element of the list of SAR values sorted in ascending order. The gradient Δ was computed for each SAR n by the following equation:

Δ_{n} = \frac{S A R_{n + 1} - S A R_{n}}{(S A R_{n} + S A R_{n + 1}) / 2},

The first significantly different value of the gradient is defined as the detection point of the outlier [62]. Figure A1 shows the voxel values corresponding to an SAR with the highest value of 0.2%, and the gradient in the TARO model.…”

Section: Figure A1mentioning

confidence: 99%

“…The removed volumes are generally located in complex shapes, such as around pinna. Note that an extensive analysis was conducted for low-frequency exposure where in situ electric fields were evaluated in voxelized models [62]. Similarly, the metric for evaluating SAR should be averaged over 10 g of tissue, and thus is not discussed extensively for RF exposures.…”

Section: Figure A1mentioning

confidence: 99%

See 1 more Smart Citation

Comparison of Thermal Response for RF Exposure in Human and Rat Models

Kodera

Hirata

2018

IJERPH

Self Cite

View full text Add to dashboard Cite

In the international guidelines/standards for human protection against electromagnetic fields, the specific absorption rate (SAR) is used as a metric for radio-frequency field exposure. For radio-frequency near-field exposure, the peak value of the SAR averaged over 10 g of tissue is treated as a surrogate of the local temperature elevation for frequencies up to 3–10 GHz. The limit of 10-g SAR is derived by extrapolating the thermal damage in animal experiments. However, no reports discussed the difference between the time constant of temperature elevation in small animals and humans for local exposure. This study computationally estimated the thermal time constants of temperature elevation in human head and rat models exposed to dipole antennas at 3–10 GHz. The peak temperature elevation in the human brain was lower than that in the rat model, mainly because of difference in depth from the scalp. Consequently, the thermal time constant of the rat brain was smaller than that of the human brain. Additionally, the thermal time constant in human skin decreased with increasing frequency, which was mainly characterized by the effective SAR volume, whereas it was almost frequency-independent in the human brain. These findings should be helpful for extrapolating animal studies to humans.

show abstract

Computational Artifacts of the In Situ Electric Field in Anatomical Models Exposed to Low-Frequency Magnetic Field

Cited by 52 publications

References 18 publications

End-to-end semantic segmentation of personalized deep brain structures for non-invasive brain stimulation

End-to-end semantic segmentation of personalized deep brain structures for non-invasive brain stimulation

An optimized block forward‐elimination and backward‐substitution algorithm for GPU accelerated ILU preconditioner in evaluating the induced electric field during transcranial magnetic stimulation

Comparison of Thermal Response for RF Exposure in Human and Rat Models

Contact Info

Product

Resources

About