Development of machine learning algorithms for the diagnosis of bovine subclinical mastitis by infrared thermography

• Grant number: 20/16240-4
• Support Opportunities: Regular Research Grants
• Start date: June 01, 2021
• End date: February 29, 2024
• Field of knowledge: Agronomical Sciences - Veterinary Medicine - Preventive Veterinary Medicine
• Principal Investigator: Luiz Francisco Zafalon
• Grantee: Luiz Francisco Zafalon
• Host Institution: Embrapa Pecuária Sudeste. Empresa Brasileira de Pesquisa Agropecuária (EMBRAPA). Ministério da Agricultura, Pecuária e Abastecimento (Brasil). São Carlos , SP, Brazil
• Associated researchers: Alexandre Rossetto Garcia
• Associated research grant(s): 24/03155-0 - 75th annual meeting of the European Animal Production Association, AR.EXT
• Associated scholarship(s): 22/04065-9 - Infrared thermography datasets for subclinical mastitis detection in dairy cattle using machine learning, BP.TT

Abstract

Mastitis is one of the major diseases in dairy cattle. The disease causes great economic losses to producers due to the reduction of milk production and changes in the quality of the product. The identification of mastitic cows is essential to reduce the number of affected mammary quarters and the duration of clinical and subclinical cases. Efforts are needed to develop accurate diagnostic techniques, which provide objective information on the health status of the mammary gland during milking and the interpretation of results without the need to send samples to laboratories, in order to streamline or improve measures of control and prophylaxis. Thus, the aim of this study is to expand information about the use of infrared thermography as an auxiliary diagnostic tool for subclinical mastitis, with the use of machine learning algorithms, enabling automation processes for improving milk quality with potential for large-scale adoption in dairy production systems. Prediction models based on these algorithms will be explored for the validation, experimentation and extraction of information from raw data sets, using information related to animals, climate, hygienic conditions of the mammary gland, milk production, infectious etiology of mastitis and status health of cows. In this way, thermographic indices using machine learning will be developed for tropical conditions, in order to increase the sensitivity, specificity and diagnostic accuracy of infrared thermography as a method for identifying mammary quarters with subclinical mastitis. Interpretations based on machine learning can be made available to make the diagnostic technique more efficient and the information that will be made available in the future can be used on cows under different types of milking, enabling decision making on the farms. (AU)