Movement analysis and recording has been confined to dedicate gait and motion capture facilities, but this only only records movements for a single part of one day in an unusual environment. In contrast medical conditions such as Parkinson’s disease and stroke, are variable and there may be key phenomena that would be difficult to collect in a gait lab. The University of Reading now has the capability to collect sensor rich data in long term studies where sensors are embedded into clothes thus enabling measurement and analysis of movements over extended periods. Initial studies have collected acceleration and gyroscope data (inertial measurements) from individuals with Parkinson’s disease and stroke and we are looking to extend these studies as well as collect data from other activities relating to health and well being. The studentship will explore these data sources and in particular will research methods of combining biomechanics and data mining techniques to identify key features of movement. The goal will be to build dynamic models of human activities such as walking, sitting, standing as well as transitions such as stand-to-sit. Techniques will include machine learning and other statistical pattern recognition ideas.
Villeneuve, E., Harwin, W., Holderbaum, W., Janko, B. and Sherratt, R. S. (2017) Reconstruction of angular kinematics from wrist-worn inertial sensor data for smart home healthcare. IEEE Access, 5. pp. 2351-2363. doi: 10.1109/ACCESS.2016.2640559