guest ::
| Home > Online First > openMINDS SANDS: making brain atlases machine- actionable using Linked Data |
| Home > Online First > openMINDS SANDS: making brain atlases machine- actionable using Linked Data |
| Poster (After Call) | FZJ-2025-04909 |
; ; ; ; ; ; ; ; ; ; ;
2025
Abstract: <b>INTRODUCTION/MOTIVATION</b><br>Brain atlases underpin many branches of neuroscience by providing a spatial scaffold on whichmultimodal data can be organised and compared. Yet atlas resources are often released as projectspecific archives differing in structure, coordinate space, parcellation logic, and nomenclature. Thisheterogeneity forces developers to write custom loaders, hinders cross-atlas comparison, andpropagates inconsistency. To overcome these barriers we introduced openMINDS SANDS(RRID:SCR_023498), a Linked Data specification that operationalises the Atlas Ontology Model(AtOM) [1] within the openMINDS metadata framework (RRID:SCR_023173). Our goal is to turnstatic atlas downloads into interoperable web resources that can be queried and reused by humansand machines alike.<br><br><b>METHODS</b><br>SANDS was implemented as an openMINDS extension integrating the four AtOM entities (referencedata, coordinate system, annotations, terminology) into the holistic openMINDS model. Schemaswere designed to follow the FAIR principles [2] (e.g., licensing, links to other FAIR resources) and toadopt AtOM suggestions (e.g., standardized atlas structures, versioning). To demonstrate expressiveness, we curated widely used brain atlases in accordance with SANDS for three commonatlas types: (i) discrete atlases where the annotation set contains only discretely defined regions, (ii)probabilistic atlases where the annotation set contains regions defined by statistically-weightedcomposites, and (iii) parcellation models where processive annotation criteria are defined to createspecimen-specific atlases by parcellating a single specimen’s anatomy and mapping it to a definedterminology. Finally, the SANDS compliant Linked Data descriptions of the curated atlases wereshared formatted as JSON-LD files through the openMINDS instance libraries (RRID:SCR_027358)making them serviceable to any software and service developers as standardized atlasrepresentations.<br><br><b>RESULTS AND DISCUSSION</b><br>As of today (2025-08-28), we have applied SANDS to 13 commonly used brain atlases andparcellation models from 3 different species. Their Linked Data representations are provided in theopenMINDS instances libraries. As evidence, we present one example for each atlas type: (i) theWaxholm Space Rat Brain Atlas (RRID:SCR_017124) [3] as example for a discrete atlas, (ii) theJulich-Brain Atlas (RRID:SCR_023277) [4] as example for a probabilistic atlas, and (iii) the DesikanKilliany Atlas [5] as example for a parcellation model. These examples demonstrate harmonized atlasusage between the EBRAINS Knowledge Graph (RRID:SCR_017612; EBRAINS central data andknowledge platform), and the siibra toolsuite [6] (EBRAINS software for providing interactivemultilevel brain atlases) enabled by adopting SANDS.By integrating AtOM into the openMINDS metadata framework, SANDS converts existing atlases intoFAIR, machine-actionable web resources. The resulting Linked Data facilitates side-by-sidevisualization, pipeline automation, and atlas-driven analysis. Moreover, SANDS instances can beharvested by search engines, enriched with community annotations, and mirrored acrossrepositories. However, the main limitation remains sociotechnical: atlas providers must supplycompliant metadata, and developers must replace hard-coded templates with dynamic queries. Toease adoption we are developing open-source converters from other standardization efforts such asBIDS [7] and provide integration support through our GitHub (Open Metadata Initiative). By makingbrain atlases first-class Linked-Data citizens, openMINDS SANDS removes the final technical barrierto fully FAIR, automation-ready neuroanatomical workflows.Keywords: Atlas Ontology Model (AtOM), discrete brain atlas, FAIR principles, Linked Data,openMINDS SANDS, parcellation model, probabilistic brain atlas, spatial reference frameworks<br><br><b>ACKNOWLEDGEMENTS</b><br>This work has received funding from the European Union’s Horizon Europe research and innovationprogramme under grant agreement No 101147319 (EBRAINS 2.0). It also was supported by theHelmholtz International BigBrain Analytics and Learning Laboratory (HIBALL).<br><br><b>REFERENCES</b><ol><li>Kleven H, Gillespie TH, Zehl L, et al. AtOM, an ontology model to standardize use of brain atlases in tools, workflows, and data infrastructures. Sci Data. 2023;10:486. https://doi.org/10.1038/s41597-023-02389-4</li><li>Wilkinson MD, Dumontier M, Aalbersberg IJJ, et al. The FAIR guiding principles for scientific datamanagement and stewardship. Sci Data. 2016;3:160018. https://doi.org/10.1038/sdata.2016.18</li><li>Kleven H, Bjerke IE, Clascá F, et al. Waxholm Space atlas of the rat brain: a 3D atlas supporting data analysis and integration. Nat Methods. 2023;20:1822-1829. https://doi.org/10.1038/s41592-023-02034-3</li><li>Amunts K, Mohlberg H, Bludau S, Zilles K. Julich-Brain: a 3D probabilistic atlas of the human brain’s cytoarchitecture. Science. 2020;369(6506):988-992. https://doi.org/10.1126/science.abb4588</li><li>Desikan RS, Ségonne F, Fischl B, et al. An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Neuroimage. 2006;31(3):968-980. https://doi.org/10.1016/j.neuroimage.2006.01.021</li><li>Dickscheid T, Gui X, Simsek A, et al. Siibra: a software tool suite for realizing a multilevel human brain atlas from complex data resources. bioRxiv. Published online May 20, 2025. https://doi.org/10.1101/2025.05.20.655042</li><li>Gorgolewski KJ, Auer T, Calhoun VD, et al. The brain imaging data structure, a format for organizing and describing outputs of neuroimaging experiments. Sci Data. 2016;3:160044. https://doi.org/10.1038/sdata.2016.44</li></ol>
|
The record appears in these collections: |
