Towards Decision Support in Digital Health Care Using Pervasive Sensing and Machine Learning

Several forms of dementia are recognized as rare diseases, such as, the spectrum of fronto-temporal degeneration and Lewy body diseases. A reported key problem in developing substantial knowledge about dementia and impacting factors is lack of data about mental processes evolving over time. Meaningful treatment of dementia today has been orientating towards multi-component interventions:  cognitive and also physical, sensorimotor stimulation appears to be promising for a meaningful treatment of dementia, however, lack of exercise is one of the major risk factors for the dementia development. Serious games for dementia are designed to motivate for playful training and pervasive sensors provide a wealth of data from interactions for the purpose of analyzing key parameters for decision support.  Objectives and results of European and national projects that focus on non-obtrusive collection and analysis of data about cognitive functions, emotional and psychosocial information are described. Various methods of machine learning were applied to extract an indicative discrimination from the automatically retrieved data about executive functions, such as, about neurodegenerative problems in inhibitory processes of attention that typically occur in rare diseases of dementia.