1. Smart Technology for Telerehabilitation: A Smart Device Inertial-sensing Method for Gait Analysis
- Author
-
Steins, D, Steins, D, Steins, D, and Steins, D
- Abstract
The aim of this work was to develop and validate an iPod Touch (4th generation) as a potential ambulatory monitoring system for clinical and non-clinical gait analysis. This thesis comprises four interrelated studies, the first overviews the current available literature on wearable accelerometry-based technology (AT) able to assess mobility-related functional activities in subjects with neurological conditions in home and community settings. The second study focuses on the detection of time-accurate and robust gait features from a single inertial measurement unit (IMU) on the lower back, establishing a reference framework in the process. The third study presents a simple step length algorithm for straight-line walking and the fourth and final study addresses the accuracy of an iPod’s inertial-sensing capabilities, more specifically, the validity of an inertial-sensing method (integrated in an iPod) to obtain time-accurate vertical lower trunk displacement measures. The systematic review revealed that present research primarily focuses on the development of accurate methods able to identify and distinguish different functional activities. While these are important aims, much of the conducted work remains in laboratory environments, with relatively little research moving from the “bench to the bedside.” This review only identified a few studies that explored AT’s potential outside of laboratory settings, indicating that clinical and real-world research significantly lags behind its engineering counterpart. In addition, AT methods are largely based on machine-learning algorithms that rely on a feature selection process. However, extracted features depend on the signal output being measured, which is seldom described. It is, therefore, difficult to determine the accuracy of AT methods without characterizing gait signals first. Furthermore, much variability exists among approaches (including the numbers of body-fixed sensors and sensor locations) to obtain useful data to
- Published
- 2016