Bradykinesia, one of the cardinal movement disorders of Parkinson’s disease manifests with different symptoms such as reduced limb velocity, amplitude, hesitancy and poverty of movement depending on the type of motor activity being performed. This variability and the fluctuations in symptom severity may explain why the autonomous tracking of bradykinesia in the context of daily activities has not been achieved. My thesis developed a software platform using advanced Machine Learning Algorithms to analyze body motion and muscle activity data to address this problem. The achievement of a 96.5% success rate for detecting bradykinesia was achieved by extracting clinically informed features from the sensor data to classify the presence of bradykinesia separately for walking and non-walking activities. A sub-set of these features were used to provide severity indicators of bradykinesia for clinical assessment.
Bhawna Shiwani is a Research fellow at Delsys Inc. She received her B.E. in Electronics and Communication Engineering from National Institute of Technology, Jalandhar, India in 2012 and she is currently pursuing her M.S. in Robotics Engineering at Worcester Polytechnic Institute, Worcester, MA. She was previously employed by Cadence Design Systems, Noida, India where she worked as a Member of Technical Staff in Analog Design Environment team. Her current research interests include artificial intelligence, control theory and cyber physical systems.