Multimodal Behavioural Analytics for Early Diagnosis of Alzheimer’s Disease

Clinical diagnosis of Alzheimer’s disease is often delayed and limited in accuracy. Although recent advances in digital health are beginning to introduce promising tools to overcome these clinical limitations, this research field is still evolving. Therefore, addressing these issues, this thesis incorporates emerging multimodal behavioural analytics, mainly focusing on changes in hand movements and gestures to identify early predictors of cognitive decline. Among the contributions of the thesis are the discovery of novel and significant indicators of early cognitive decline, methodological approaches and dataset for future research. These contributions assist in developing digital tools with improved accuracy and sensitivity facilitating early diagnosis of Alzheimer's disease.