PETsuch

PETsuch is a term used to describe an open, interoperable framework intended to facilitate discovery and reuse of positron emission tomography (PET) imaging data across medical research institutions. It combines a standardized metadata catalog with a search and discovery layer, enabling researchers to locate datasets by tracer, disease, anatomical region, scanner model, acquisition protocol, or study design. The goal is to accelerate multi-center analyses while preserving patient privacy.

The concept emerged from collaborations among academic hospitals, imaging centers, and data science consortia in the late 2010s. A reference implementation and data model were proposed to align with existing standards such as DICOM and ontology mappings. The project is typically described as a community effort with open-source tooling and ongoing governance, oriented toward enabling reproducible research and broader data sharing within ethical and legal frameworks.

Architecture and methods include a metadata registry, a search index, and a secure data-access layer. PETsuch uses ontologies to map free-text queries to structured fields, enabling semantic search over complex imaging datasets. Privacy-preserving techniques, de-identification, and consent management are integral, and portal access supports role-based permissions. Interoperability is achieved through standardized data models and API-based integrations with hospital information systems and imaging archives, facilitating cross-institution queries without exposing protected information.

Impact and challenges center on enabling meta-analyses, cohort discovery for trial recruitment, and retrospective studies across centers. Key hurdles include data heterogeneity, incomplete metadata, varying consent terms, and governance across institutions. Success depends on community standards, sustained funding, and ongoing collaboration among stakeholders in imaging informatics.