Congratulations to NetwellCASALA’s Andrea Kealy who passed her PhD viva on December 15th. Andrea’s thesis is titled ‘Use of Ambient Sensors for the Longitudinal Profiling of the Nocturnal Behaviour of Older Adults in their Homes’. Andrea was supervised by Dr. John Loane and Dr. Kevin McDaid.
Andrea’s thesis abstract:
With an ageing population putting an increased demand on healthcare, an alternative means for employing regular monitoring is required in order to maintain the quality of life of older people as they age. Research indicates that patient-centric monitoring is the ideal means for health management. Remote monitoring using a series of ambient sensors has been proposed as a means for supporting the monitoring required for a patient-centric healthcare model. Facilitating such monitoring could aid in the identification of early signs of age related illnesses, and as a consequence, support early intervention.
This thesis focuses on the investigation and development of an unobtrusive method to monitor day and night time behaviour using an ambient sensor setup. The long-term deployment of ambient sensors within the homes of the older adults living in the Great Northern Haven serves as the testbed for this research. The novel contributions contained within this thesis focus on the development and validation of models to extract day and night time behaviour measures. This incorporates the development of a validation framework, encompassing four tiers of validation, proposed to evaluate the reliability of the derived behaviour measures. The work presented extends current research through the profiling of the short and medium term behaviour of older adults living in their own homes. Furthermore the short term change in the behaviour of older adults surrounding significant events is investigated to explore whether a change in day and night time behaviour could provide important information on the behaviour of adults immediately before and after illness events.
While changes in sleep patterns normally occur as people age degradation in sleep over time has been shown to have a considerable negative impact on health. It is envisaged that the novel use of density map visualisations, representing movement in three main living areas of individuals’ homes, could facilitate the identification of changes in night time behaviour. Additionally the application of cluster analysis to automatically classify changes in night time behaviour is explored.