Toolboxes for SEEG Electrode Localization and Visualization
Abstract number :
3.252
Submission category :
5. Neuro Imaging / 5A. Structural Imaging
Year :
2022
Submission ID :
2205095
Source :
www.aesnet.org
Presentation date :
12/5/2022 12:00:00 PM
Published date :
Nov 22, 2022, 05:28 AM
Authors :
Armin Vosoughi, MD – Emory University; David Brandman, MD, PhD – Assistant Professor, Neurosurgery, University of California, Davis; Ammar Kheder, MD – Neurology – Emory University; Leonardo Bonilha, MD, PhD – Neurology – Emory University; Adam Dickey, MD – Neurology – Emory University; Daniel Drane, PhD – Neurology – Emory University; David Gutman, MD, PhD – Neurology – Emory University; Nigel Pedersen, MD – Neurology – Emory University
Rationale: Drug-resistant patients with focal epilepsy often require intracranial monitoring to delineate the surgical target while sparing the eloquent cortex. In particular, the method of stereoelectroencephalography (SEEG) requires precise localization and visualization of the depth electrodes contacts to correctly interpret the clinical data. There are many non-commercial (i.e., open-source) toolboxes available with this very goal. However, they provide various features and are tailored for specific use cases. The ideal toolbox would have (1) a built-in multi-modality co-registration, (2) automatic contact localization, (3) procedures to correlate electrophysiologic information with contacts, (4) atlas-based localization, (5) 3D visualization of contacts within a custom atlas and (6) group analysis. We analyzed each platform for these six characteristics.
Methods: We reviewed free or open-source toolboxes that aim to localize and visualize surface and depth electrodes and their individual contacts. Given our transition from subdural grids to SEEG using depth electrodes about five years ago, we focused on toolboxes amenable to depth electrode data. We reviewed the dependencies and features of each platform. We then evaluate each toolbox in ease of use and features for SEEG practitioners and researchers.
Results: Our search revealed 18 functional free toolboxes for research purposes. Most toolboxes need either FreeSurfer or SPM for segmentation and coregistration purposes, however some depend also on FSL and ITK-SNAP. Tools vary in their requirements for image registration. Most of the toolboxes were based on MATLAB (i.e., SEEGview, iELVIS, iEEGview, iElectrodes, IntraNat Electrodes, Ntools, moviEEG, FieldTrip, eConnectome, BrainMapper, GARDEL, LeGUI), some on Python (IELU, img_pipe, 3dSlicer, MNE, BrainQuake), and one on C++ (BioImage Suite). Many used either Visualization Toolkist (VTK, Kitware Inc.) or Insight Toolkit (ITK) toolboxes for visualization, along with Mayavi, and MATLAB runtime. Most of the tools were compatible with both Linux and Mac OS X, and many were also executable in Windows. None of the toolboxes met all six criteria, to our knowledge. However, MNE, BrainQuake, and LeGUI included most of the ideal criteria. Of the 18 toolboxes we reviewed, only 4 (i.e., LeGUI, BrainQuake, elec_ntools, and MNE) were actively maintained for use in SEEG electrode localization with code commits in the last 12 months.
Conclusions: Localizing intracranial electrodes and visualizing each contact within the brain against standardized atlases is vital for research and clinical purposes. Finding an ideal toolbox that can carry out all the desired functions is challenging, as most toolboxes are built for specific purposes that could not be used globally. In addition, some of the toolboxes that have been introduced in the past years depend on outdated dependencies and are not compatible with most of the current configurations, especially on Linux and Mac.
Funding: NPP was supported by CURE Epilepsy, NIH R21NS122011, and K08NS105929.
Neuro Imaging