Machine Learning-Based Analysis of Human Motions for Parkinson’s Disease Diagnostics

Body language is known to be a powerful medium for carrying information about the mental and physical states of a human being. In medicine, visual observations were used by physicians to base the diagnosis on diseases that affect human motor functions. In neurology, different motor tests have been used for more than a century to diagnose cognitive and neurodegenerative disorders such as Parkinson’s disease or Alzheimer’s disease. Technological advances in motion acquisition techniques have made it possible to register human movement parameters in a way that is invisible to the naked eye. Coupled with advances in machine learning, this has sparked studies that have focused on supporting diagnostics on the basis of human motion analysis. The talk delivered in the colloquium summarises the research of our workgroup in this area.