Authors :
Nishant Sinha, PhD – University of Pennsylvania
Dominique Duncan, PhD – University of Pennsylvania
Darrell DeFreitas, - – University of Pennsylvania
Jacqueline Boccanfuso, MRA – University of Pennsylvania
Vijay Samayamantula, MS – University of Pennsylvania
Kathryn Davis, MD – Center for Neuroengineering and Therapeutics and Penn Epilepsy Center, Department of Neurology, University of Pennsylvania
Brian Litt, MD – University of Pennsylvania
Presenting Author: Joost Wagenaar, PhD – University of Pennsylvania
Rationale:
Epilepsy affects over 70 million people worldwide and more than 3.6 million Americans, one-third of whom do not respond to medication. While new technologies, implantable devices, high-bandwidth intracranial EEG (iEEG), and multimodal imaging have improved treatments, meaningful data sharing is lacking, which limits clinical progress and translational research. There is a tremendous need to aggregate, share, and collaborate on increasingly large and complex multimodal data to advance these efforts. The Epilepsy.Science platform (https://epilepsy.science) has been funded to provide a scalable solution for large-scale data integration, analysis, and sharing in an impactful way. Our group has focused on building tools and a community to accelerate translational epilepsy and neuroscience research, including the IEEG-Portal (IEEG.org). Independently, the Pennsieve platform was developed over the last 7 years as an open-source, cloud-based data management and sharing platform for large volumes of multimodal data in the neurosciences, with a focus on standards, scalability, and sustainability. It is currently used as the data core for several NIH programs (e.g., SPARC and REJOIN) and is ideally suited to replace iEEG.org as the next-generation repository for multimodal data in epilepsy research.
Methods:
Epilepsy.Science offers services to find, access, interrogate, and reuse public Epilepsy Datasets uploaded to the Pennsieve platform and our partner platforms. By standardizing datasets across repositories, Epilepsy.Science provides mechanisms to support cohort creation, and cross-validation of analysis tools for biomarker discovery. All data from IEEG.org will be curated into a standardized BIDS format, and metadata will be structured according to the OMOP Common Data Model. The platform supports advanced mechanisms to visualize EEG data directly in the browser and provide services to run analysis pipelines over data on the platform without downloading the raw data files.
Results:
Currently, over 2000 scientific datasets are being curated and made available on Epilepsy.Science in a standardized manner. Our partnerships with other repositories will ensure that Epilepsy.Science will serve as a single-entry point for accessing data across various epilepsy-related repositories. Epilepsy.Science will provide tools and resources to easily find and leverage common data elements associated with Epilepsy Research. This initiative will enhance collaboration among the repositories to improve data availability and accessibility through shared protocols.
Conclusions:
In conclusion, Epilepsy.Science is a newly developed resource built on over a decade of experience fostering collaborative research for the epilepsy research community. As the quantity and complexity of data have surged over the last decade, it is increasingly important to provide scalable solutions for data integration and analysis that meet the needs of the community. This includes ensuring datasets are AI-ready, that analyses can be run next to the data in the cloud, and that the data platform is developed with scalability and sustainability in mind.
Funding: This effort was funded by the NIH: 5U24NS134536 and 3U24NS134536