Use of sensor technology to determine health status in cattle and sheep

The overall objective of this thesis was to quantify the ability of new technologies to identify the health status of cattle and sheep, and to identify the factors which affect the diagnostic capabilities of these technologies. The PhD consisted of four experiments which investigated: i) the repeatability of infrared thermography in an agricultural environment, ii) the relationship between udder skin temperature (measured by infrared thermography) and the quarter somatic cell count of dairy cows, and iii) the ability of infrared thermography to detect lameness in sheep, and iv) the use of a custom hoof weigh crate to detect lameness in sheep. Approximately 8000 thermal images were manually captured throughout the PhD in an agricultural setting; these data were analysed using a variety of statistical methods including variance component analysis, mixed models, random regression models, and sensitivity and specificity analyses. Results from this thesis showed that to gain the required level of repeatability from thermal image measurements three replicate images must be captured by a trained operator. As part of this thesis, the effect of the environment on udder skin temperature was quantified but thermal imaging could not be used to estimate somatic cell count. The maximum temperature of sheep hooves proved to be a useful variable to diagnose early onset of infection in individual hooves. This thesis was the first to suggest that infrared thermography used in agriculture has the best diagnostic capabilities at colder ambient temperatures. The custom hoof weigh crate developed as part of this thesis was capable of measuring the individual hoof load of sheep, but was only a viable solution for detecting extensive infection rather than the early onset of lameness. Results from this thesis will aid in the application of infrared thermography in real world conditions, especially as a lameness detection tool for sheep