Data-based modeling of recovery from mastitis in dairy cows

The prevention, detection and treatment of mastitis forms the basis for the contemporary udder-health management. After detection and startup of treatment, recovery; being the suppression of the causal pathogen (bacteriological healing), the disappearance of clinical symptoms and the regeneration of the udder tissue (clinical healing) isn't objetively monitored. Due to lack of information about this recovery, it is impossible for the dairy farmer to estimate the effectiveness of the treatment and to adjust the duration of treatment and further support. This results in excessive antibiotic use in the case of a too long treatment and, on the other hand, the absence of complete recovery by too short or suboptimal treatment.

In previously conducted studies, two models were developed to estimate and monitor the severity of the infection and the inflammatory response in mastitis, as well as the recovery:

1) A production model that makes it possible to accurately calculate the milk losses by mastitis at quarterly level

2) A recovery model based on (historical) animal data and dairy production- & quality records currently available on modern dairy farms.

In this project, these models are optimized and validated on Flemish dairy farms so that this knowledge can be used to improve the udder-health.