DABI stimulates progress in the neurosciences by organizing, storing, and disseminating human invasive neurophysiology data along with tools for analysis and visualization. Investigators retain ownership and control of their data through a federated model. Building on decades of experience creating widely-used, large-scale informatics solutions in the neurosciences, the team behind DABI has launched a new archive to ingest, harmonize, aggregate, store, visualize, and disseminate human invasive neurophysiology data, including EEG, ECoG, LFP, single unit activity, and more. The repository, which will also house synchronized behavioral, imaging, demographic, and other key data, is specifically designed to help BRAIN Initiative researchers organize and analyze their own data while fulfilling data-sharing directives from federal agencies and their respective institutions.
THE PLATFORM, DATA TYPES, AND INFRASTRUCTURE
The DABI platform is designed to ingest, harmonize, aggregate, store, visualize, and disseminate human invasive neurophysiology data from BRAIN Initiative investigations. It provides a seamless interface for researchers uploading data and novel, intuitive ways for users to search and analyze the archive’s contents. In addition to comprehensive search, download, and data upload functionality, as well as streamlined data upload capabilities, DABI includes an interactive data visualization environment and integration with a suite of data-processing tools. DABI accommodates a wide variety of invasive neurophysiology data types, as well as imaging, behavioral and clinical data, in order to create a comprehensive picture of the human brain in health and disease. The infrastructure supporting DABI is comprised of fail-safe, redundant and secure components, more than 7PB of storage capacity and significant compute power.
Data Partners have a variety of study information and use different codes and naming conventions.
Metadata exported in CSV files are mapped into a common model comprised of common database schema, data dictionary and code lists.
An ON/OFF switch gives Data Partners full control over their data. When the switch is off data is completely removed from the DABI network.
Data can be cached to computer memory but is never written to DABI server disks.
Investigators query and visualize combined data to gain insight about invasive neurophysiology.
INTERACTIVE DATA VISUALIZATION
Visualizing data is a key way to orient investigators, enabling the discovery of hidden insights and unexamined relationships between attributes. Thus, the DABI team has created a comprehensive interface to facilitate data visualization within the platform itself. Investigators can view data correlations and trends, including correlations between biomarkers and clinical outcomes, within downloading and important data into an external application.
LONI DATA-PROCESSING PIPELINE
The LONI Pipeline Processing Environment—an extensible framework of protocols for spatial normalization, template mapping, atlas construction, statistical mapping, and format conversion—has been used to analyze human neuroscience data for hundreds of research publications. Researchers can use the tool to easily access imaging and neurophysiology data and process it in parallel within DABI. The LONI Pipeline team is continually developing innovative software to enhance the environment and users can upload their own analytic tools if desired.