Engineering Vet-OncoNet: expert-governed harmonisation and joint topography–morphology classification of veterinary pathology data

Routine veterinary pathology data can support cancer surveillance, but differences in laboratory terminology and data formats hinder their integration. We developed and implemented a Vet-OncoNet framework that preserves submitted data while harmonising records from seven independent pathology laboratories. Expert-curated dictionaries standardise demographic and other categorical variables. For oncological classification, the framework interprets reported topography and diagnosis jointly, because anatomical location alone may not identify the tissue from which a lesion arose. Previously validated pairs are resolved through a curated memory; unfamiliar pairs undergo computational candidate assessment, with uncertain or conflicting decisions referred for expert review. In the evaluated production cycle, 58,510 records entered the Veterinary International Classification of Diseases for Oncology (Vet-ICD-O) stage. Expert assessment was requested for 3,897 normalised topography–diagnosis pairs representing 7,894 records. Twelve pairs, corresponding to 19 records, were excluded, leaving 58,491 uniquely traceable records. Of these, 50,616 (86.5%) followed an automatic classification pathway: 45,474 through reuse of curated pair decisions and 5,142 through the computational pathway. Harmonised data supported provider-specific dashboards, which evolved from seven separate Data Studio outputs to a single R Shiny application with laboratory-specific access. Vet-OncoNet shows how expert-defined terminology, joint topography–morphology classification and selective computational support can be combined in an adaptable surveillance workflow. This implementation provides a framework that other animal cancer registries can adapt to harmonise routine pathology data while retaining traceability and expert oversight.

Authors

Institutions

Publication Details

Journal
Frontiers in Veterinary Science
Published
2026-11-18
DOI
https://doi.org/10.3389/fvets.2026.2027454
Primary Topic
Veterinary Oncology Research
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Engineering Vet-OncoNet: expert-governed harmonisation and joint topography–morphology classification of veterinary pathology data

João Niza‐Ribeiro, Katia Cristina Pinello
Frontiers in Veterinary Science
Veterinary Oncology Research
article

Engineering Vet-OncoNet: expert-governed harmonisation and joint topography–morphology classification of veterinary pathology data

João Niza‐Ribeiro, Katia Cristina Pinello
article en

Abstract

Routine veterinary pathology data can support cancer surveillance, but differences in laboratory terminology and data formats hinder their integration. We developed and implemented a Vet-OncoNet framework that preserves submitted data while harmonising records from seven independent pathology laboratories. Expert-curated dictionaries standardise demographic and other categorical variables. For oncological classification, the framework interprets reported topography and diagnosis jointly, because anatomical location alone may not identify the tissue from which a lesion arose. Previously validated pairs are resolved through a curated memory; unfamiliar pairs undergo computational candidate assessment, with uncertain or conflicting decisions referred for expert review. In the evaluated production cycle, 58,510 records entered the Veterinary International Classification of Diseases for Oncology (Vet-ICD-O) stage. Expert assessment was requested for 3,897 normalised topography–diagnosis pairs representing 7,894 records. Twelve pairs, corresponding to 19 records, were excluded, leaving 58,491 uniquely traceable records. Of these, 50,616 (86.5%) followed an automatic classification pathway: 45,474 through reuse of curated pair decisions and 5,142 through the computational pathway. Harmonised data supported provider-specific dashboards, which evolved from seven separate Data Studio outputs to a single R Shiny application with laboratory-specific access. Vet-OncoNet shows how expert-defined terminology, joint topography–morphology classification and selective computational support can be combined in an adaptable surveillance workflow. This implementation provides a framework that other animal cancer registries can adapt to harmonise routine pathology data while retaining traceability and expert oversight.

Frontiers in Veterinary ScienceVol. 13
Universidade do Porto (PT), Instituto de Ciências Biomédicas Albel Salazar (PT), Institute of Public Health, University of Porto (PT)
Veterinary Oncology Research
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon