The Computational Movement Science (CMS) Lab is a research group of the Movement Science department of the Institute of Sport and Exercise Science. Its focus lies on model-based analyses of various aspects of human movement and perception. Our key idea is to combine internal forward and inverse models into one unified artificial neural network that controls and predicts movement in a parallel fashion. We are continuously developing a computational 3D human model, called Myonardo, with bones, muscles, ligaments, reflexes, and motor control units. Myonardo is based on human physiology and can be adapted to fit any individual in any kind of circumstance, using kinetic and kinematic measurements, electromyography, and state of the art neuronal networks implemented in MATLAB and Simulink. As a result, motion tracking and mixed dynamics analyses with Myonardo enable a detailed evaluation of the data. We use artificial neural networks, such as reservoir computing networks and Kohonen maps. The data is evaluated with advanced statistical methods, such as general and generalized linear models. We investigate how motor skills are learned and improved, and how actual and phantom pain emerges. Closely related basic theories comprise the reafference principle, internal models in the brain, eye-hand coordination and balancing.
Current research topics:
- Developing artificial neural networks to control myo-electrical hand prosthetics, for the purpose of using EMG patterns to precisely predict hand and finger movements for the application to hand prosthetics Frederik Thies
- Predicting joint forces based on whole-body kinetics with Myonardo, and comparing these predictions with the experimental data of the CAMS-knee project Heiko Wagner & Kim Boström
- Calculating ground reaction forces using kinetic and kinematic data and implemation of the Ghent footmodel in Myonardo Kim Boström, Adrian Honc & Rosemary Dubbeldam & Pauline Laufer
- Recreating the human body as a 3D-Model by researching imaging technology and designing a workflow for implementing Photogrammetry into Myonardo visualisation Oliver Schmitz
- Reservoir computing in Motor Control: Motor Switching and Motor Variability Myriam de Graaf
- Myonardo development, Motor Control and Motor Memory Kim Boström
- Interactive movement optimization Adrian Honc
- Visualization with the 3D animation platform Unity Frederic Thies, Oliver Schmitz
Ongoing cooperative work:
• Vrije Universiteit Amsterdam
• Karl-Franzens University Graz
• Friedrich-Schiller-Universität Jena.