Juliana Klein
Juliana Klein
Advisors
Dr. Markus Zrenner, Mathias Maurer, Burkhard Duemler (adidas AG), Dr. Stefan Sesselmann, Prof. Dr. Björn Eskofier
Duration
09 / 2017 – 02 / 2018
Abstract
A stable knee is of great importance in sport and everyday life. Lack of stability can cause injuries like ligament ruptures and may result in the end of a professional’s career. Based on literature research several risk factors for an unstable knee are presented in this work and a test battery is composed to increase and analyze stabilization. This test battery consists of several exercises including balance, strength and jump tests. As there is not always a physiotherapist nearby to determine the knee stability, there is a demand for a training aid to get an automated feedback. To this end, an existing sensor equipped knee sleeve has been used. It contains two inertial sensors recording acceleration and angular velocity data which were analyzed for predefined parameters to give feedback to the user.
In this master thesis algorithms based on a subsequence dynamic time warping (SDTW) approach will be developed. The algorithms extract and evaluate parameters, such as jump height and event detection, based on data acquired by the knee sleeve. A study with 16 healthy subjects has been conducted to evaluate the accuracy of the developed algorithms on the test battery. As reference systems a force plate and also a motion capturing system were used. The jump height was calculated using the flight-time-method and hereby compared to the force plate data. The mean absolute error for drop jumps is 2.1 cm, squat jumps 4.1 cm and for single leg squat jumps 3.5 cm (right) and 3.7 cm (left). The sensitivity for detecting glute bridges and squats ranges from 96.6-98.6% while the sensitivity for single leg hops is 91.0-99.2%. Furthermore, biomechanical experts evaluated videos of the performed jumps and rated the take-off and landing phases. The expectations were met and showed a correlation between qualitative and quantitative results.