Digital reconstruction of microscopic anatomies, a technique employed by the neuroscientist community for decades, allows for robust qualitative and quantitative analysis. However, it is common for output files of existing segmentation tools to be absent of detailed metadata about the subject, sample, and anatomies. Use of cross-species anatomical ontologies (e.g. Uberon) have not been widely applied in image analysis and neuroanatomical reconstruction pipelines because no tools have existed that serve anatomical terms from the ontologies directly to the user at the time of segmentation.
To overcome this challenge, we have established comprehensive pipelines for findable, accessible, interoperable, and reusable (FAIR) mapping of multi-species, multi-organ, and multi-scale anatomical regions by developing tools that not only support the use of ontologies at the time of data creation, but also define each data element to enable the repurposing or remixing of segmentation data for other scientific inquiries.
The components of this toolset include direct API connection to the SciCrunch knowledge infrastructure within MBF Bioscience software for delivery of current, FAIR, and organ-specific terminology lists curated by anatomical experts; verification methods to check that all annotations have globally unique and persistent identifiers; and a specification of our neuromorphological file format which richly characterizes all elements of data produced within this toolset. Together, these novel integrations and specifications allow researchers to reconstruct, analyze, and share meaningful and metadata-enriched digital representations of multiscale anatomical data.