Deep learning can automate chicken tibia-breaking strength quantification to improve animal welfare
Poultry Science · 2026
DOI: 10.1016/j.psj.2026.106549 ↗Kurz gefasst
Fachartikel in der Zeitschrift Poultry Science, 2026. Verfasst von Tanmay Debnath und 7 weiteren Autor:innen. Beteiligte Einrichtung: The Roslin Institute, University of Edinburgh, Easter Bush, Midlothian EH25 9RG, UK. Electronic address: T.Debnath@sms.ed.ac.uk. Hauptthemen laut MEDLINE: Chickens, Tibia, Animal Welfare, Deep Learning. Die Arbeit nennt 49 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Foundation for Food and Agriculture Research, University of Edinburgh, Biotechnology and Biological Sciences Research Council.
Abstract (Originalsprache)
Bone damage is an important welfare issue in the poultry industry, yet large-scale phenotyping of chicken bone strength currently relies on time-consuming manual annotation of X-rays or destructive post-mortem testing. To address this, an end-to-end deep-learning pipeline was developed that automatically (i) segments the chicken tibiotarsus from lateral X-ray images (U-Net, Dice = 0.91) and (ii) predicts its breaking strength from pixel intensities alone. Using 916 curated bone images, the predictor achieved moderately high correlation with measured breaking strength (maximum Pearson's correlation of 0.74), exceeding the performance of a previous labour-intensive manual annotation method. Image-derived predictions were moderately heritable (h² ≈ 0.16) and exhibited an exceptionally high genetic correlation with the physical trait, indicating that selection on the model-derived phenotype is a good proxy to select for bone strength. The workflow therefore provides a potential rapid, non-invasive and genetically informative alternative to post-mortem testing, paving the way for the routine incorporation of bone-quality traits into commercial breeding programmes and improved poultry welfare at scale.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- The Roslin Institute, University of Edinburgh, Easter Bush, Midlothian EH25 9RG, UK. Electronic address: T.Debnath@sms.ed.ac.uk.
- The Roslin Institute, University of Edinburgh, Easter Bush, Midlothian EH25 9RG, UK. Electronic address: peter.wilson@roslin.ed.ac.uk.
- The Roslin Institute, University of Edinburgh, Easter Bush, Midlothian EH25 9RG, UK. Electronic address: ricardo.pong-wong@roslin.ed.ac.uk.
- The Roslin Institute, University of Edinburgh, Easter Bush, Midlothian EH25 9RG, UK. Electronic address: Lindsey.Plenderleith@ed.ac.uk.
- Lohmann Breeders GmbH, 27472 Cuxhaven, Germany. Electronic address: B.Andersson@lohmann-breeders.com.
- Lohmann Breeders GmbH, 27472 Cuxhaven, Germany. Electronic address: M.Schmutz@lohmann-breeders.com.
Förderung
- Foundation for Food and Agriculture Research
- University of Edinburgh
- Biotechnology and Biological Sciences Research Council
Zitieren
APA
Debnath, T., Wilson, P., Pong-Wong, R., Plenderleith, L., Andersson, B., Schmutz, M., Dunn, I., & Prendergast, J. G. D. (2026). Deep learning can automate chicken tibia-breaking strength quantification to improve animal welfare. Poultry Science. https://doi.org/10.1016/j.psj.2026.106549
Vancouver
Debnath T, Wilson P, Pong-Wong R, Plenderleith L, Andersson B, Schmutz M, et al. Deep learning can automate chicken tibia-breaking strength quantification to improve animal welfare. Poultry Science. 2026. doi:10.1016/j.psj.2026.106549
BibTeX
@article{Debnath2026,
author = {Debnath, Tanmay and Wilson, Peter and Pong-Wong, Ricardo and Plenderleith, Lindsey and Andersson, Björn and Schmutz, Matthias and Dunn, Ian and Prendergast, James G D},
title = {Deep learning can automate chicken tibia-breaking strength quantification to improve animal welfare},
journal = {Poultry Science},
year = {2026},
doi = {10.1016/j.psj.2026.106549}
}
EUVIMED