Sensor-Based Gait Analysis Using Biomechanical Modeling and Optimal Control (LME 1290)

Art der Arbeit:

Master Thesis

Institution:

Lehrstuhl für Informatik 5 (Mustererkennung)

Betreuer:

Dorschky, Eva
Lehrstuhl für Maschinelles Lernen und Datenanalytik
Telefon +49 9131 85 27890, E-Mail: eva.dorschky@fau.de

Hannink, Julius
Lehrstuhl für Maschinelles Lernen und Datenanalytik
Telefon +49 9131 85 27921, E-Mail: julius.hannink@fau.de

Heiko Schlarb (adidas AG)

Klucken, Jochen

Prof. Dr. Antonie van den Bogert (Cleveland State University)

Leyendecker, Sigrid
Lehrstuhl für Technische Dynamik (LTD, Prof. Leyendecker)
Telefon 09131/85-61001, Fax 09131/85-61011, E-Mail: sigrid.leyendecker@fau.de

Eskofier, Björn
Lehrstuhl für Maschinelles Lernen und Datenanalytik
E-Mail: bjoern.eskofier@fau.de

Beschreibung der Arbeit:

Sensor-based gait analysis can give valuable insight into human health. For example, gait analysis with two foot-mounted inertial sensors is used in order to assist in the diagnosis and therapeutic monitoring of Parkinson's disease [1]. One major challenge is to achieve a high quality gait analysis based on noisy and sparse sensor measurements. Moreover, inertial sensors can only quantify human joint kinematics and are not able to measure joint kinetics as performed in gait laboratories. Existing systems are based on an integration of the inertial sensor data for estimating human poses [2]. This error-prone integration could be avoided using a computer simulation of a biomechanical model that tracks the measured sensor signals. Furthermore, such a model could give insight into joint kinetics, muscle control and other gait-related parameters such as stride length, stride time and ground-reaction force. The purpose of this thesis is the evaluation of a method for sensor-based gait analysis using an optimal control simulation of a biomechanical model [4].
This includes the following tasks:
- Literature research on:

• Inertial sensor based human motion tracking: What commercial systems, patents and publications exist? What are the differences and drawbacks?

• Enhancing motion capturing by optimal control simulation of human movement
  

- Implementation of:

• Simulated accelerometer and gyroscope signals for a given musculoskeletal model [3]

• Optimal control simulation of human gait based on tracking multiple sensor axes [4]
  

- Evaluation:

• Validation study with at least 10 subjects using optical marker system as gold standard

• Comparison of simulated sensor signals to actual sensor measurements

• Comparison of simulated kinematics and kinetics to optical motion capture system

• Evaluation for tracking a reduced number of sensor axes

• Optional: Comparison of simulated gait-related parameters to optical motion capture system

Implementations of the algorithms must be provided in Matlab and C during the design and evaluation phases. The implementation has to follow the lab guidelines and complete source code and documentation have to be provided. The thesis must contain a detailed description of the implementation as well as a profound result evaluation and discussion. Additionally, a profound research on literature, existing patents and related work in the corresponding working areas have to be performed.
References
[1] J. Barth et al. (2011). Biometric and mobile gait analysis for early diagnosis and therapy monitoring in Parkinson's disease. EMBC, pp. 868-871.
[2] J. Zhang et al. (2013). Concurrent validation of xsens MVN measurement of lower limb joint angular kinematics. Physiol Meas, 38(4), pp. 63-69.
[3] A. J. van den Bogert et al. (1996). A method for inverse dynamic analysis using accelerom- etry. J Biomech, 29(7), pp. 949-954.
[4] A. J. van den Bogert et al. (2011). Implicit methods for eficient musculoskeletal simulation and optimal control. IUTAM, 2, pp. 297-316.

Bearbeitungszustand:

Die Arbeit ist bereits abgeschlossen.
Bearbeiter:	Iris Kellermann
Abgegeben am:	03.05.2016