AUTHOR=Sagrera Mònica , Cobos Àlex , Garza-Moreno Laura , Pérez Mónica , García-Buendía Gema , Huerta Eva , Llorens Anna Maria , Espigares David , Sibila Marina , Segalés Joaquim 

TITLE=Automated pixel-based quantification of porcine circovirus 2 genome in formalin-fixed, paraffin-embedded tissues using in situ hybridisation

JOURNAL=Frontiers in Veterinary Science

VOLUME=Volume 12 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/veterinary-science/articles/10.3389/fvets.2025.1609897

DOI=10.3389/fvets.2025.1609897

ISSN=2297-1769

ABSTRACT=IntroductionDetection of porcine circovirus 2 (PCV2) in lymphoid tissues is essential for diagnostic and research purposes. In situ hybridisation (ISH) enables the localisation of viral genomes in tissue sections but is traditionally assessed visually, which may introduce subjectivity.MethodsThis study developed an automated pixel classifier to quantify the PCV2 genome using RNAscope® ISH technology. Four lymphoid tissues (tonsils and tracheobronchial, mesenteric, and superficial inguinal lymph nodes) from 66 experimentally inoculated pigs were analysed. PCV2 labelling was assessed both visually (scores 0–3) and digitally (percentage of labelled area).ResultsA strong correlation was observed between visual and digital ISH scoring (ρ = 0.96), allowing the definition of digital thresholds corresponding to visual scores. Among all tissues, TBLN exhibited the highest PCV2 labelling. This tissue was further evaluated by PCV2 quantitative polymerase chain reaction (qPCR), showing a high correlation with digital ISH results (ρ = 0.85).DiscussionThese findings demonstrate the reliability of digital pathology tools for objective quantification of PCV2 in lymphoid tissues. Automated scoring enhances consistency, reduces observer bias, and improves diagnostic efficiency in PCV2 research and surveillance.