Extraction of Information from Inertial Sensors to Aid Health Assessment
Dandu, Sriram Raju, Electrical Engineering - School of Engineering and Applied Science, University of Virginia
Lach, John, Department of Electrical and Computer Engineering, University of Virginia
Tracking human activity and fitness have been the latest trends in the world of wearable technology. Inertial sensors offer a practical and low cost method for objectively monitoring human movements, and particularly have the applicability to monitor free-living subjects. Shoes, smart phones, watches and other wearables equipped with inertial sensors are able to capture human motion in a non-obtrusive way. Similar sensors are placed on several positions on the human body to monitor a range of different movements, including gait, sit-to-stand transfers, postural sway and limb movements. Inertial sensors have provided a simple and reliable approach to collect human kinematics data. The inertial data collected by these sensors contains valuable information related to human movements, which can be used to monitor and assess the health of a person. But the field of wearable technology still struggles to develop a methodology which can derive relevant information, from the raw inertial data, to aid in health assessment.
This work presents a framework to extract information from raw inertial data corresponding to human motion. The framework is a sequential learning process coupled with application-specific knowledge. This framework has three stages: data preparation, feature engineering and information extraction. The first stage converts the raw data into usable data, the second stage extracts relevant attributes from the data and the third stage derives information from these attributes to aid health assessment. A number of techniques can be used at each of the stages, and the appropriate ones are chosen by employing application-specific knowledge as the cornerstone.
I will be implementing the framework on two real world case studies; one studies the relationship between inertial features and gait pathology in Multiple Sclerosis and the other one deals with classification of head impacts in contact sports using inertial sensors. The information derived from both the case studies has the potential to serve as a tool to monitor the health of a person.
MS (Master of Science)
Wearables, Feature Extraction, Pattern Recognition, Accelerometer, Machine Learning, Multiple Sclerosis, Head Impact Classification, Dynamic Time Warping, Linear Dynamical Model, Modelling, Health Assessment, Inertial Sensors