EUVIMEDEUVIMED entdecken
Crossref

312 Leveraging digital phenotyping from research to implementation: Insights from a pig breeding company

Ching-Yi Chen, Eric Psota, Bruno Valente, Pieter Knap, Justin W Holl

Journal of Animal Science · 14.09.2024 · 102 (Supplement_3) · 50-51

DOI: 10.1093/jas/skae234.055 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Animal Science, online erschienen am 13.09.2024. Verfasst von Ching-Yi Chen und 4 weiteren Autor:innen. Die Arbeit wurde bisher 4-mal zitiert (Crossref).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In pig breeding, the integration of phenotypes into genetic improvement has been employed for decades. With today’s advancements in modern sensors, digital phenotyping provides a promising avenue for further enhancing accuracy through genomic selection. This integration also creates opportunities to study areas that used to be challenging due to limitations in data collection. We present four examples of digital phenotypes to illustrate how a company may leverage digital phenotyping for genetic implementation: individual feed intake, computer vision leg scoring, behavior, and metabolomics. Phenotypes recorded using automated electronic devices such as feed intake devices enable continuous tracking and recording of individual feed intake events when animals enter the feeder station during performance testing. The full series of events is often condensed into a single average per individual prior to genetic evaluation. By implementing computer vision, the precision of leg scoring phenotypes has increased from subjective, human-measured categorical scores of 1 to 9 to objective, camera-based continuous assessments. We have implemented convolutional neural networks (CNNs) to predict leg scores in our selection process. Using automation, the camera-based leg scoring system enables us to extend the scale of data used in genetic evaluations to improve animal welfare and longevity. Digital phenotyping holds promise for behavioral studies. Traditional methods of phenotype collection for individual behavior have been hindered by challenges such as the difficulty of capturing individual identification and the time-consuming nature of data collection. We have developed a proprietary, custom ear tag reading method and tracking algorithm that enables the automated identification of unique pigs even in visually challenging conditions. The method estimates behavior traits including time spent eating, drinking, lying, sitting, standing, and meters of distance traveled. These are used for genetic parameter estimation with investigations on association with production traits. Understanding metabolic profiles of pigs offers the potential to uncover new traits that were previously difficult to measure, and to identify biomarkers associated with traits. In conclusion, the integration of digital phenotyping into pig breeding programs offers a path toward more precise and efficient genetic improvement. Digital data capture yields a large amount of data which still provides novel opportunities for research and application.

Zitieren

APA

Chen, C., Psota, E., Valente, B., Knap, P., & Holl, J. W. (2024). 312 Leveraging digital phenotyping from research to implementation: Insights from a pig breeding company. Journal of Animal Science, 102(Supplement_3), 50-51. https://doi.org/10.1093/jas/skae234.055

Vancouver

Chen C, Psota E, Valente B, Knap P, Holl JW. 312 Leveraging digital phenotyping from research to implementation: Insights from a pig breeding company. Journal of Animal Science. 2024;102(Supplement_3):50-51. doi:10.1093/jas/skae234.055

BibTeX
@article{Chen2024,
  author = {Chen, Ching-Yi and Psota, Eric and Valente, Bruno and Knap, Pieter and Holl, Justin W},
  title = {312 Leveraging digital phenotyping from research to implementation: Insights from a pig breeding company},
  journal = {Journal of Animal Science},
  year = {2024},
  volume = {102},
  number = {Supplement_3},
  pages = {50--51},
  doi = {10.1093/jas/skae234.055}
}
Kostenlos starten →