TemplateFlow: FAIR-sharing of multi-scale, multi-species brain models.

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Brain Atlasing

Background:

• Standardized neuroimaging analysis relies on reference brain anatomies (templates) and atlases.
• Current methods for sharing these essential resources are fragmented, often bundled with software or available through informal downloads.
• Lack of a centralized registry hinders resource accessibility and reproducibility.

Purpose of the Study:

• Introduce TemplateFlow, a novel, publicly available framework for managing and distributing human and non-human brain models.
• Establish a standardized system for sharing brain templates and atlases.
• Promote FAIR data principles (Findable, Accessible, Interoperable, Reusable) for neuroimaging resources.

Main Methods:

• Development of an open-access database for brain templates and atlases.
• Creation of associated software for resource access, management, and quality control (vetting).
• Implementation of a framework supporting resource sharing under FAIR principles.

Main Results:

• TemplateFlow provides a unified platform for accessing diverse brain models across species.
• The framework facilitates "multiverse analyses" to test result generalizability across different references and scales.
• Ensures efficient and standardized access to crucial neuroimaging resources.

Conclusions:

• TemplateFlow addresses the critical need for a centralized, FAIR-compliant registry of brain templates and atlases.
• Enhances the reproducibility and standardization of neuroimaging research.
• Supports cross-species comparative studies and robust validation of neuroimaging findings.