Andmebaasi logo

Sirvi Autor "Banhazi, T.M." järgi

Nüüd näidatakse 1 - 1 1
  • Pisipilt
    Kirje
    Development and pilot testing of the ‘Lesion-Detect’ system for automated identification and scoring of gross lung lesions in pigs during slaughter
    (Estonian University of Life Sciences, 2026) Banhazi, T.M.; Dunn, M.; Banhazi, A.; Reixach, J.; Gol, S.; Ramon-Perez, A.; Llonch, P.; Vujacic, V.; Sudar, J.; Kajari, N.; Maselyne, J.; Van Noten, N.; Tuyttens, F.
    Respiratory diseases are the most significant health and welfare challenges in pig production. Enzootic pneumonia (EP) is often subclinical at farm level, making early detection difficult, yet its economic and welfare impacts are substantial. Suboptimal housing conditions are well-established risk factors that predispose pigs to respiratory disease. These conditions negatively affect animal health and welfare and are associated with increased mortality, higher medication costs, reduced feed intake, sub-optimal average daily gain, and increased organ condemnations, representing a major constraint to sustainable pig production. At slaughter, gross lung lesions associated with respiratory disease can be assessed in all carcasses, offering a valuable opportunity for herd-level monitoring. However, current Pig Health Monitoring System (PHMS) lesion scoring methods are labour-intensive, subjective, and poorly scalable. To address these limitations, we developed and pilot-tested Lesion-Detect (LD), an automated image-analysis system for identifying and scoring gross lung lesions under commercial slaughterhouse conditions. LD prototypes were deployed in two European slaughterhouses (Spain: hanging lungs; Serbia: tray-based presentation) and benchmarked against manual PHMS scoring. The system showed strong agreement with trained human assessors, with false positive and false negative rates typically within 2–3% of lung area, comparable to inter-observer variability. By delivering objective and scalable lesion scoring, LD could enable consistent respiratory health surveillance. Beyond data collection, rapid herd-level feedback supports earlier and more targeted on-farm interventions, potentially reducing antimicrobial use and improving animal welfare. With further refinement, the system may be extended to additional carcass-level welfare indicators, strengthening its role within precision livestock farming and integrated animal welfare assessment frameworks.