|dc.description.abstract||Disease is widely acknowledged to be detrimental to an animal’s welfare and generates significant economic losses. To minimise the impact of disease on animal welfare and consequently reduce economic costs, early disease detection is essential for promoting earlier diagnosis, treatment and isolation to prevent the spread of disease. As automation increases in the livestock industries, there is a need to develop and integrate early disease indicators into automated systems for remote monitoring of animal health and welfare. The aims of this thesis were to firstly, investigate the behavioural (milk feeding, lying and water drinking) and physiological (infrared temperatures and respiration rate) responses which occur during the onset of neonatal calf diarrhoea (NCD), and to assess their suitability as early disease indicators. Additionally, due to an increasing trend towards providing calves with greater milk allowances, the influence of milk allowance on indicator suitability was assessed. Secondly, to determine the optimum anatomical location (eye, cheek, side, shoulder and back) for infrared image collection for early disease detection. Thirdly, to investigate the water intake and drinking behaviour of calves and how these behaviours are affected by disease onset and milk allowance. Further, to develop and validate an algorithm for the automated detection and analysis of the eye and cheek regions from infrared images and further validate the use of infrared thermography as an automated method of recording respiration rate (RR) in young calves. Finally, the thesis aimed to develop and validate an automated water system for the purpose of recording drinking behaviour and further assessed the suitability of setting up an infrared camera at a water station for monitoring calf health and welfare.
Calves were defined as being clinically ill when they were observed as being diarrhoeic. Milk feeding, lying and water drinking behaviours changed prior to clinical signs of disease. Changes in milk feeding and lying behaviours prior to clinical signs were thought to occur due to calves experiencing a reduction in appetite, becoming lethargic and attempting to conserve energy in response to the onset of disease. Infrared temperatures of the eye, cheek and shoulder decreased and side (situated over the rumen fossa) temperature increased prior to clinical signs. The decrease in infrared temperatures was attributed to the animal generating a fever to fight against the infection by restricting blood flow to the extremities, in order to maintain homeostasis. Further, the decrease in infrared temperature was likely a result of a reduction in feed and metabolic activity during the onset of disease. The side was identified as the optimum anatomical location for early disease detection. The increase in side temperature was attributed to the proximity of the area to the underlying site of infection and localised inflammation of the intestines. Prior to clinical signs, there was little change in water drinking behaviour (number and duration of visits to the water trough), which was considered to be due to the young age of the calves at the time of diagnosis. Additionally, a lack of a response in RR prior to clinical signs was thought to be due to sampling rate and NCD being an enteric disease as opposed to a respiratory disease.
Milk allowance influenced the suitability of behavioural and physiological measures as early disease indicators. When provided with either a high (10 L/d) or low (5 L/d) milk allowance, milk feeding behaviours typically only changed for calves on a high milk allowance and infrared temperatures (eye and cheek) only changed for calves on a low milk allowance. Additionally, as opposed to being used individually, combining behavioural and physiological measures provided stronger composite indicators of disease. Combinations of milk feeding and lying behaviours, and additionally infrared temperatures (eye and cheek) in the case of calves on a low milk allowance, provided the strongest composite indicators of disease.
Integrated into automated systems, behavioural and physiological measures could be utilised for early disease detection. In support of the future development of integrating infrared thermography (IRT) into automated systems, an algorithm was developed and validated for detection and analysis of the eye and cheek regions from infrared images. Further, the use of IRT as a method of recording respiration rate was validated for use in young calves. In addition, an automated water system was validated for the purpose of recording drinking behaviour (number and duration of visits to the water trough). Based on the findings that calves visited the water trough on average three times per day and began consuming water during the first week of life a water system was considered as a suitable and more affordable alternative to an automated calf feeder for setting up an infrared system to monitor calf health and welfare.
Although these measures have been assessed in relation to the detection of NCD, they have the potential to be applied further for the detection of other diseases and in other species. The integration of these measures into an automated on-farm system could enable earlier treatment and isolation of diseased animals to prevent the spread of disease, by alerting farmers to diseased animals earlier than is currently possible based on overt clinical signs. In addition to facilitating decision-making abilities for the farmer, the development of such a system would reduce costs on-farm and to the industry as a whole and would ultimately improve animal health and welfare.||