The detection of health alterations in dairy cattle using 3D-vision based reticulo-ruminal motility assessment in a milking robot

Abstract

Background: The reticulo-ruminal contraction frequency (RCF) comprises the contractions of the rumen to redistribute the ingesta. A well-functioning reticulo-rumen is important for dairy cattle to maintain a good health. An alteration in RCF often appears in diseased cows. Continuous information over time is useful to monitor the health of dairy cattle. Therefore, an automated 3D-vision based RCF-assessment was developed. The research objective of this study was to investigate whether there is an association between the automated RCF-assessment and health alterations in dairy cows, using clinical examination. Materials and methods: 50 cows on a Dutch dairy farm were examined 18 times during 9 weeks. An examination protocol was created to obtain a clinical score (CS) that represents the health alterations in cows. The reticulo-rumen was recorded during every milking session in a milking robot by a 3D-camera. After data processing, the RCF was obtained and could be compared with the CS. Results: No significant association has been found between health alterations and RCF. However, individual cases of sick cows reveal that there are changes in RCF, which could indicate a possible association in sick cows. Furthermore, there is a slight association between RCF and milk production and ration change. Discussion: The 3D-vision based RCF-assessment measures the RCF accurate and is cheap, non-invasive and causes no distress in cows. Future longitudinal studies with this 3D-vision based assessment in a milking robot are therefore recommended to further investigate how this tool can contribute to the monitoring of the cow’s health. The ultimate goal is to join the available data of all sensors together to improve the disease detection of the milking robot.