Identifier: TFM:206
Authors: Batista de Frutos, Edgar
Abstract:
Study of Human Wandering: An ICT approach AUTHOR: Edgar Batista de Frutos DIRECTOR: Dr. Agusti Solanas Abstract Society is experiencing a steady ageing process due to the growth of the population and the increase of life expectancy. Consequently, the number of age-related diseases, such as mild cognitive impairments (MCI) and dementia, are gaining increasing importance. People suffering from this kind of illnesses notice a slight deterioration in their cognitive abilities, such as forgetting events, and they might also start experiencing behavioural disorders. A common and problematic behavioural disorder is wandering, which leads to the appearance of temporal episodes of disorientation and confusion. In this context, the early detection of wandering episodes would help to ensure the safety of wanderers and to enhance the quality of life of their relatives and caregivers. Wandering detection is not a trivial task due to the multiple factors that affect it, such as the daily patterns of the wanderer, the degree of severity of the illness or the kind of environment. With the widespread use of smartphones and ubiquitous technology, new healthcare services aiming at reducing the impact of wandering have been proposed. Yet, the research community have not been able to propose a universal method to automatically detect wandering behaviour, independently of the wanderer or the environment. In this dissertation, we propose two new methods to automatically detect wandering by analysing trajectories. Both methods are validated and compared with the most relevant methods in the literature, using real trajectories from online datasets, together with elderly trajectories gathered during a research project conducted in the Smart Health Research Group. Education: Master’s Degree in Computer Engineering: Computer Security and Intelligent Systems Presentation Date: February 2017
Repositori URV