Measurement of Posture Using Advanced Sensor Technologies - A Literature Study and Trial Experiment of an IMU-Based System

Abstract

Abstract The purpose of this thesis was to identify a quantitative postural assessment tool for monitoring the progress of rehabilitation therapies, as current visual methods are subjective and prone to errors. The goal was to measure sagittal angles along the entire spine to provide an objective indicator of postural balance. A literature review was conducted to provide an overview of conventional clinical methods and to explore novel postural assessment tools available on the market. The most promising emerging methods included Kinect, inertial measurement unit (IMU) systems, and smartphone applications. While Kinect and smartphone-based methods offered limited accuracy and seemed more appropriate for estimating gross misalignments, IMU systems demonstrated superior precision, versatility, and applicability to both dynamic and static measurements, though at a higher cost. Based on these findings, the Movella DOT IMU system was selected for evaluation in a trial experiment. Four sensors were placed along the spine of a test subject, and results were compared with the Qualisys motion capture system during sitting, standing, and walking tasks. The sensor placed at the cervical lordosis showed the strongest agreement, with an overall mean bias of 1.6°. When gait data were excluded, the cervical placement yielded a mean bias of 0.41° and a mean RMSD of 0.64°, which is considered an excellent result. Sensors placed at lower spinal regions exhibited slightly higher deviations. Further development of dedicated software and validation in larger test groups is necessary for clinical implementation but is considered highly feasible. The Movella DOT system has the potential to serve as an all-in-one solution for multiple needs in rehabilitation therapy, combining accuracy with versatility, and thus is a highly promising tool for clinical postural assessment.