Abstract
The use of wearable technology has grown rapidly in recent years with the development of low-cost sensors and accelerometers. Their use in tracking behavioural activity has gained popularity as an affordable measure of health and well-being in humans. These health applications are also being applied to wearable accessories for animals, including pets. Therefore, this study aimed to use a triaxial accelerometer to evaluate the behavioural patterns in household dogs. First, the typical activity patterns of healthy household dogs, their chronotype and correlation with their owner and other dogs in the household were determined. This established a ‘key time of the day’ when dogs were least influenced by their owner and thus more likely to be performing their own behaviours.
By establishing when the dog is most likely to be performing their own behaviours, corresponding to when the correlation between owner’s routine and dog routine is minimal, this could aid the creation of predictive biomarkers in canine health. If significant changes in dog behaviour were to occur at these times, the changes would more likely be due to a change in ill health instead of a direct effect of a change in the owner’s routine. Then this research explored behavioural indicators key to different chronic canine illnesses; osteoarthritis, cognitive dysfunction and pruritus. The symptomatic behaviours were examined, as well as other behaviours: head shaking, eating, drinking, immobile and Quality of Life questionnaire data, to get a holistic view of behavioural patterns impacted through different conditions. Following this, the research showed the accelerometer could detect different behavioural patterns unique to each condition. Moreover, these behavioural changes were different at this established ‘key time of day’. Overall, this emphasises the benefits of looking at combinations of behaviours over time as predictive behavioural biomarkers for changes in health and welfare.