[Paper Review] Development of an Ontology for an Integrated Image Analysis Platform to enable Global Sharing of Microscopy Imaging Data

Satoshi Kume, Hiroshi Masuya+2 more|arXiv (Cornell University)|Oct 20, 2021

Biomedical Text Mining and Ontologies3 references9 citations

TL;DR

This paper proposes an RDF/OWL-based ontology extending the Open Microscopy Environment (OME) data model to integrate optical and electron microscopy imaging data with biosample, bioresource, and experimental metadata. By translating OME's XML schema into RDF and adding 18 upper-level concepts—including electron microscopy, phenotype data, and imaging conditions—it enables global, interoperable sharing and integrated analysis of microscopy data across modalities and databases.

ABSTRACT

Imaging data is one of the most important fundamentals in the current life sciences. We aimed to construct an ontology to describe imaging metadata as a data schema of the integrated database for optical and electron microscopy images combined with various bio-entities. To realise this, we applied Resource Description Framework (RDF) to an Open Microscopy Environment (OME) data model, which is the de facto standard to describe optical microscopy images and experimental data. We translated the XML-based OME metadata into the base concept of RDF schema as a trial of developing microscopy ontology. In this ontology, we propose 18 upper-level concepts including missing concepts in OME such as electron microscopy, phenotype data, biosample, and imaging conditions.

Motivation & Objective

• To develop a standardized, interoperable metadata schema for global sharing of microscopy imaging data across optical and electron microscopy modalities.
• To address the lack of integrated ontologies for biosamples, bioresources, and experimental conditions in existing microscopy data models.
• To extend the OME data model with RDF/OWL to support semantic integration with external biological and medical databases.
• To enable precise linking of microscopy images to detailed biological information, such as strain and sample data, via standardized ontologies.
• To support large-scale, reproducible image analysis by structuring metadata for computational and multidisciplinary integration.

Proposed method

• Translated the OME data model (v. January 2015, XML-based) into an RDF/OWL schema to establish a standardized data model for imaging metadata.
• Identified and extracted core concepts and properties from the OME XML schema to define RDF classes and relationships.
• Extended the OME ontology with 18 new upper-level concepts, including Image, SampleContainer, BioSample, PhenotypeData, ImagingCondition, and ElectronMicroscopyDevice.
• Integrated biosample and bioresource metadata by introducing dedicated classes and linking them to external databases such as the RIKEN BioResource Center.
• Defined RDF graph instances to demonstrate how microscopy images can be semantically linked to experimental conditions, sample preparations, and phenotype data.
• Validated the ontology by describing approximately 20,000 electron microscopy datasets with diverse staining methods, confirming its scalability and expressiveness.

Experimental results

Research questions

• RQ1 How can an ontology be designed to unify metadata from optical and electron microscopy across diverse experimental conditions?
• RQ2 What additional upper-level concepts are required to extend the OME data model to support electron microscopy and biosample integration?
• RQ3 Can an RDF/OWL-based ontology enable interoperable, machine-readable integration of microscopy data with external biological databases?
• RQ4 To what extent can the extended ontology support large-scale, reproducible image analysis and data sharing across institutions?
• RQ5 How can imaging metadata be semantically linked to phenotype data and bioresources for enhanced data reusability?

Key findings

• The proposed ontology successfully extends the OME data model with 18 new upper-level concepts, including ElectronMicroscopyDevice, PhenotypeData, and ImagingCondition, enabling comprehensive metadata description.
• The RDF/OWL schema supports the integration of electron microscopy data with biosample and bioresource metadata, bridging gaps in current imaging data models.
• The ontology enables precise semantic linking of microscopy images to external databases, such as the RIKEN BioResource Center, via standardized entity references.
• The framework was validated on approximately 20,000 electron microscopy datasets with diverse staining methods, demonstrating scalability and practical applicability.
• The extended ontology supports multidisciplinary data integration and enables global sharing of microscopy data through alignment with linked open data principles.
• The approach facilitates future integration with Biological Dynamics Markup Language (BDML) and Cellular Microscopy Phenotype Ontology (CMPO), enhancing cross-domain compatibility.