Dataset Information

Predicting in vivo MRI Gradient-Field Induced Voltage Levels on Implanted Deep Brain Stimulation Systems Using Neural Networks.

ABSTRACT: Introduction:MRI gradient-fields may induce extrinsic voltage between electrodes and conductive neurostimulator enclosure of implanted deep brain stimulation (DBS) systems, and may cause unintended stimulation and/or malfunction. Electromagnetic (EM) simulations using detailed anatomical human models, therapy implant trajectories, and gradient coil models can be used to calculate clinically relevant induced voltage levels. Incorporating additional anatomical human models into the EM simulation library can help to achieve more clinically relevant and accurate induced voltage levels, however, adding new anatomical human models and developing implant trajectories is time-consuming, expensive and not always feasible. Methods:MRI gradient-field induced voltage levels are simulated in six adult human anatomical models, along clinically relevant DBS implant trajectories to generate the dataset. Predictive artificial neural network (ANN) regression models are trained on the simulated dataset. Leave-one-out cross validation is performed to assess the performance of ANN regressors and quantify model prediction errors. Results:More than 180,000 unique gradient-induced voltage levels are simulated. ANN algorithm with two fully connected layers is selected due to its superior generalizability compared to support vector machine and tree-based algorithms in this particular application. The ANN regression model is capable of producing thousands of gradient-induced voltage predictions in less than a second with mean-squared-error less than 200 mV. Conclusion:We have integrated machine learning (ML) with computational modeling and simulations and developed an accurate predictive model to determine MRI gradient-field induced voltage levels on implanted DBS systems.

SUBMITTER: Erturk MA

PROVIDER: S-EPMC7044348 | biostudies-literature | 2020

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Predicting <i>in vivo</i> MRI Gradient-Field Induced Voltage Levels on Implanted Deep Brain Stimulation Systems Using Neural Networks.

Erturk M Arcan MA Panken Eric E Conroy Mark J MJ Edmonson Jonathan J Kramer Jeff J Chatterton Jacob J Banerjee S Riki SR

Frontiers in human neuroscience 20200220

<h4>Introduction</h4>MRI gradient-fields may induce extrinsic voltage between electrodes and conductive neurostimulator enclosure of implanted deep brain stimulation (DBS) systems, and may cause unintended stimulation and/or malfunction. Electromagnetic (EM) simulations using detailed anatomical human models, therapy implant trajectories, and gradient coil models can be used to calculate clinically relevant induced voltage levels. Incorporating additional anatomical human models into the EM simu ...[more]

PMID: 32153375

Similar Datasets

Project description:BackgroundDeep brain stimulation (DBS) is an established treatment for patients with movement disorders. Patients receiving chronic DBS provide a unique opportunity to explore the underlying mechanisms of DBS using functional MRI. It has been shown that the main safety concern with MRI in these patients is heating at the electrode tips - which can be minimised with strict adherence to a supervised acquisition protocol using a head-transmit/receive coil at 1.5T. MRI using the body-transmit coil with a multi-channel receive head coil has a number of potential advantages including an improved signal-to-noise ratio.Study outlineWe compared the safety of cranial MRI in an in vitro model of bilateral DBS using both head-transmit and body-transmit coils. We performed fibre-optic thermometry at a Medtronic ActivaPC device and Medtronic 3389 electrodes during turbo-spin echo (TSE) MRI using both coil arrangements at 1.5T and 3T, in addition to gradient-echo echo-planar fMRI exposure at 1.5T. Finally, we investigated the effect of transmit-coil choice on DBS stimulus delivery during MRI.ResultsTemperature increases were consistently largest at the electrode tips. Changing from head- to body-transmit coil significantly increased the electrode temperature elevation during TSE scans with scanner-reported head SAR 0.2W/kg from 0.45°C to 0.79°C (p<0.001) at 1.5T, and from 1.25°C to 1.44°C (p<0.001) at 3T. The position of the phantom relative to the body coil significantly impacted on electrode heating at 1.5T; however, the greatest heating observed in any position tested remained <1°C at this field strength.ConclusionsWe conclude that (1) with our specific hardware and SAR-limited protocol, body-transmit cranial MRI at 1.5T does not produce heating exceeding international guidelines, even in cases of poorly positioned patients, (2) cranial MRI at 3T can readily produce heating exceeding international guidelines, (3) patients with ActivaPC Medtronic systems are safe to be recruited to future fMRI experiments performed under the specific conditions defined by our protocol, with no likelihood of confound by inappropriate stimulus delivery.

Project description:ObjectiveTo correct for image distortions produced by standard Fourier reconstruction techniques on low field permanent magnet MRI systems with strong [Formula: see text] inhomogeneity and gradient field nonlinearities.Materials and methodsConventional image distortion correction algorithms require accurate [Formula: see text] maps which are not possible to acquire directly when the [Formula: see text] inhomogeneities also produce significant image distortions. Here we use a readout gradient time-shift in a TSE sequence to encode the [Formula: see text] field inhomogeneities in the k-space signals. Using a non-shifted and a shifted acquisition as input, [Formula: see text] maps and images were reconstructed in an iterative manner. In each iteration, [Formula: see text] maps were reconstructed from the phase difference using Tikhonov regularization, while images were reconstructed using either conjugate phase reconstruction (CPR) or model-based (MB) image reconstruction, taking the reconstructed field map into account. MB reconstructions were, furthermore, combined with compressed sensing (CS) to show the flexibility of this approach towards undersampling. These methods were compared to the standard fast Fourier transform (FFT) image reconstruction approach in simulations and measurements. Distortions due to gradient nonlinearities were corrected in CPR and MB using simulated gradient maps.ResultsSimulation results show that for moderate field inhomogeneities and gradient nonlinearities, [Formula: see text] maps and images reconstructed using iterative CPR result in comparable quality to that for iterative MB reconstructions. However, for stronger inhomogeneities, iterative MB reconstruction outperforms iterative CPR in terms of signal intensity correction. Combining MB with CS, similar image and [Formula: see text] map quality can be obtained without a scan time penalty. These findings were confirmed by experimental results.DiscussionIn case of [Formula: see text] inhomogeneities in the order of kHz, iterative MB reconstructions can help to improve both image quality and [Formula: see text] map estimation.

Dataset Information

Predicting in vivo MRI Gradient-Field Induced Voltage Levels on Implanted Deep Brain Stimulation Systems Using Neural Networks.

Publications

Predicting <i>in vivo</i> MRI Gradient-Field Induced Voltage Levels on Implanted Deep Brain Stimulation Systems Using Neural Networks.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets