Enhancing collaborative neuroimaging research: introducing COINSTAC Vaults for federated analysis and reproducibility

Area of Science:

• Neuroscience
• Bioinformatics
• Data Science

Background:

• Collaborative neuroimaging research faces significant barriers, including technological, policy, and administrative challenges, despite ample data availability.
• Existing platforms often require data sharing, raising privacy and logistical concerns.
• Federated analysis offers a solution by enabling analysis without direct data sharing.

Purpose of the Study:

• To introduce COINSTAC Vaults (CVs) as a novel enhancement to the COINSTAC platform.
• To reduce barriers in collaborative neuroimaging research by providing standardized, persistent, and highly-available datasets.
• To integrate CVs with COINSTAC's federated analysis capabilities for streamlined, self-service research.

Main Methods:

• Development and integration of COINSTAC Vaults (CVs) into the existing COINSTAC platform.
• Implementation of a user-friendly interface for self-service analysis.
• Utilizing CVs for hosting standardized datasets, including open-access data, for federated analysis.

Main Results:

• CVs successfully host standardized, persistent, and highly-available datasets.
• Seamless integration with COINSTAC's federated analysis capabilities was achieved.
• Demonstrated improved collaboration and reduced manual coordination through self-service analysis.
• Showcased potential for increased research reproducibility and larger sample sizes in neuroimaging studies.

Conclusions:

• COINSTAC Vaults significantly reduce barriers in collaborative neuroimaging research.
• CVs enhance data accessibility and streamline the federated analysis process.
• The platform facilitates reproducible research and expands opportunities for larger-scale neuroimaging studies.