SUNConferences, 27th Annual Conference of the SA Institute for Industrial Engineering

Font Size: 
Development of an mHealth rehabilitation activity monitoring system for transtibial amputees
Talon Garikayi, Dawie van den Heever, Stephen Matope

Last modified: 2016-10-06

Abstract


This paper presents a novel approach for a remote activity and mobility rehabilitation monitoring system for transitibial amputees. The proposed system has the capability of monitoring ambulatory and mobility related activities such as standing, walking, sitting and walking on stairs. The design utilizes single channel inertial measurement signals to determine amputee performance. The methods involved in the system design include development of an embedded data acquisition system, data storage system, signal processing algorithm and the user interface. The final module of the system includes a data processing and presentation algorithm for clinical applications. Design parameters were based on normal human gait. The hardware and software for context-aware ubiquitous computing applications was developed. Subjects were tasked to perform random activities for three hours and the data was collected and analyzed. The system classified motions as walking, standing, laying, sitting and walking on stairs. The classification accuracy was 93.3%, due to challenges on classifying between walking upstairs and walking downstairs. The system error was 6.7%.


Keywords


Healthcare Systems, Biomechatronics, Artificial Intelligence , mHealth