Predictive models for heat stress assessment in Holstein dairy heifers using infrared thermography and machine learning.

Autor: Pereira ALV; Department of Biosystems Engineering, Faculty of Animal Science and Food Engineering (FZEA), University of São Paulo (USP), Av. Duque de Caxias Norte 225. Campus Fernando Costa, Pirassununga, SP, 13635-900, Brazil. alvp.11.11@gmail.com., Martello LS; Department of Biosystems Engineering, Faculty of Animal Science and Food Engineering (FZEA), University of São Paulo (USP), Av. Duque de Caxias Norte 225. Campus Fernando Costa, Pirassununga, SP, 13635-900, Brazil., Campos JCD; Department of Biosystems Engineering, Faculty of Animal Science and Food Engineering (FZEA), University of São Paulo (USP), Av. Duque de Caxias Norte 225. Campus Fernando Costa, Pirassununga, SP, 13635-900, Brazil., da Silva Rodrigues AV; Department of Biosystems Engineering, Faculty of Animal Science and Food Engineering (FZEA), University of São Paulo (USP), Av. Duque de Caxias Norte 225. Campus Fernando Costa, Pirassununga, SP, 13635-900, Brazil., de Carvalho Nunes Oliveira GP; Department of Biosystems Engineering, Faculty of Animal Science and Food Engineering (FZEA), University of São Paulo (USP), Av. Duque de Caxias Norte 225. Campus Fernando Costa, Pirassununga, SP, 13635-900, Brazil., de Sousa RV; Department of Biosystems Engineering, Faculty of Animal Science and Food Engineering (FZEA), University of São Paulo (USP), Av. Duque de Caxias Norte 225. Campus Fernando Costa, Pirassununga, SP, 13635-900, Brazil.
Jazyk: angličtina
Zdroj: Tropical animal health and production [Trop Anim Health Prod] 2024 Sep 26; Vol. 56 (8), pp. 285. Date of Electronic Publication: 2024 Sep 26.
DOI: 10.1007/s11250-024-04152-2
Abstrakt: Heat stress is a condition that impairs the animal's productive and reproductive performance, and can be monitored by physiological and environmental variables, including body surface temperature, through infrared thermography. The objective of this work is to develop computational models for classification of heat stress from respiratory rate variable in dairy cattle using infrared thermography. The database used for the construction of the models was obtained from 10 weaned heifers, housed in a climate chamber with temperature control, and submitted to thermal comfort and heat wave treatments. Physiological and environmental data were collected, as well as thermographic images. The machine learning modeling environment used was IBM Watson, IBM's cognitive computing services platform, which has several data processing and mining tools. Classifier models for heat stress were evaluated using the confusion matrix metrics and compared to the traditional method based on Temperature and Humidity Index. The best accuracy obtained for classification of the heat stress level was 86.8%, which is comparable to previous works. The authors conclude that it was possible to develop accurate and practical models for real-time monitoring of dairy cattle heat stress.
(© 2024. The Author(s), under exclusive licence to Springer Nature B.V.)
Databáze: MEDLINE
Externí odkaz: