Abstracts

Rationale: Patients who undergo implantation of intracranial electrodes to detect seizure onset zones (SOZ) for drug-resistant epilepsy undergo a workup of that consists of multimodal data. This data includes structural and functional imaging (MRI, CT and PET), electrical stimulation mapping (ESM) and electrophysiological analysis results. To best make use of these numerous data types and the unique information and insights each provides it becomes convenient to have the ability to visualize them all at once. Seecog provides an intuitive and easy to use interface to interact with and visualize multimodal data in single and multiple subjects that undergo implantation of intracranial electrodes.

Methods: SEECOG is built using Javascript and is rendered using additional HTML and CSS files. It makes use of existing Javascript libraries such as AMI.js, Tabulator, jquery, jquery-ui and Three.js. It’s built to integrate directly with a iELVis, a Matlab-based electrode localization visualization toolbox built around other freely available software. However, SEECOG can use data produced by other toolboxes as well as Fieldtrip, MNE-Python, and Bioimagesuite. Imaging data and electrodes are rendered using AMI.js and THREE.js. Electrode appearance and attributes can be further customized and adjusted in an interactive table.

Results: This browser-based tool can be run either offline by opening an html file or visiting its website hosted on Github. No additional installation of software is required. Data loaded into Seecog is kept locally and not uploaded to a remote server this allowing data containing PHI to be used. It allows visualization in a 3D viewer and orthogonal planes of volumetric data, surface reconstructions of cortical and subcortical regions, intracranial electrodes, parcellations and functional overlays from fMRI and PET. The accompanying table allows easy single-click annotation of electrode properties such as SOZ, interictal spiking and if the electrode elicited any response during ESM related to functions such as motor or language. Electrode appearance can be customized based on these annotations by changing their color, shape and size. Additionally, the table provides databasing capabilities such as filtering, querying and categorizing electrodes. In addition, data from multiple subjects can be loaded into Seecog and rendered on a template brain such as fsaverage.

Conclusions: This software and its updates compliments previously existing applications by providing a friendly and intuitive user-interface to interact with multimodal data from epilepsy patients that is accessible to all users regardless of technical background. Compared to a version of this application presented at the AES meeting 2020, this version includes new modules such as electrodes from multiple subjects loaded at once, parcellations, snapping electrodes to surfaces, Python and Matlab scripts to format data for Seecog, bug fixes, a smoother user interface, and other changes. Researchers, in addition to clinicians, can hopefully benefit from Seecog when interpreting research results across subjects by allowing viewing of multimodal data, categorizing and querying of electrodes across research subjects.

Funding: Not applicable