Development of Algorithms for Computing Knee Stability Parameters Using a Sensor Equipped Knee Sleeve

Klein J, Zrenner M, Dümler B, Eskofier B (2019)

Publication Type: Conference contribution, Abstract of a poster

Publication year: 2019

Event location: Weiden DE

Abstract

A stable knee is of great importance as lack of stability can cause injuries like ligament ruptures and may result in the end of a professional athlete’s career. Several studies confirm the effectiveness of prevention trainings for knee stabilization and can reduce the risk for anterior cruciate ligament ruptures up to 90 % [1]. This paper describes a new approach to measure parameters of knee stabilizing exercises using a sensor equipped knee sleeve [2].

Based on literature research several risk factors for an unstable knee were identified (e.g. weak knee flexors, proprioception deficits) and a test battery with stabilizing exercises was composed. The test battery included balance, strength and jump tests. Based on those exercises, we developed algorithms to evaluate the knee stability using data from a knee sleeve equipped with two inertial measurement units (IMUs) located on the thigh and the shank. The algorithms using a subsequence dynamic time warping (sDTW) approach are capable of computing jump height using a flight-time-approach and the number of repetitions of strength exercises.

We conducted a study with 16 subjects to evaluate the accuracy of the developed algorithms on the test battery. As gold standard, we used a force plate for computing jump height and videos to count the number of repetitions. The jump height computed using the data obtained by the knee sleeve showed an error of 3.3 cm. The sensitivity of the sDTW algorithm used to extract the repetitions in the IMU data streams was 97.6 %.

The results show that the sensor equipped knee sleeve can accurately extract knee stability relevant
parameters. These parameters can be used to monitor and improve the knee stability using training plans, which can not only prevent ligament ruptures, but also help in recovery scenarios.

Authors with CRIS profile

Markus Zrenner Lehrstuhl für Informatik 14 (Machine Learning and Data Analytics) Björn Eskofier Lehrstuhl für Informatik 14 (Machine Learning and Data Analytics)

Additional Organisation(s)

Machine Learning and Data Analytics Lab

Involved external institutions

Adidas AG DE Germany (DE)

How to cite

APA:

Klein, J., Zrenner, M., Dümler, B., & Eskofier, B. (2019, June). Development of Algorithms for Computing Knee Stability Parameters Using a Sensor Equipped Knee Sleeve. Poster presentation at ORTHO – Kongress der Master, Orthopädie 4.0, Weiden, DE.

MLA:

Klein, Juliana, et al. "Development of Algorithms for Computing Knee Stability Parameters Using a Sensor Equipped Knee Sleeve." Presented at ORTHO – Kongress der Master, Orthopädie 4.0, Weiden 2019.

BibTeX: Download