Keynote Lectures

Abstract

Sports are becoming more and more technological. Engineering in general and Information Technology, Artificial Intelligence, Computer Vision and Simulation in particular are becoming an important and almost essential support for many activities directly or indirectly related to sport sciences with emphasis on high performance sports training. In this growing world of high performance sports, elite coaches and sport stars are increasingly entering into the new innovation and technological world to advance performance and gain competitive advantages. This talk will analyze the current state of the art on Sports Analysis and High Performance Training and overview recent projects on these areas. The talk will also present research work and recent advances developed at our laboratory in the area of collective sports analysis, player models and team models creation and extraction and collective sports simulation. Our work has been targeting on creating high-level realistic simulators based on realistic player models in order to enable elite coaches to test and fine tune game strategies for distinct sports both indoor and outdoor.

Abstract
Coordination of the body segments is one of the most fundamental skills facilitating human motion. Only through successful coordination of the movements of all body segments it is possible to create, to transfer, and to counterbalance the forces that are necessary for executing a movement task. This talk will discuss principal component analysis (PCA) as one method to quantify and investigate coordination patterns in human motion. Effective application of PCA depends on an understanding of the various sources that create variability in human motion data. The talk will therefore also discuss such sources and suggest methods that can be used to reduce detrimental effects of unwanted variability in the analysis. Various application examples for PCA will be presented including clinical gait research, performance analysis in sports, or sports engineering research.

Abstract
The lecture will be focused on the characterization muscle-tendon mechanical properties using elastography. A brief review will present elastographic methods and indicate advantages of each. Since it provides an instantaneous measurement, an ultrasound method based on the shear wave velocity measurement called Supersonic Shear Imaging (SSI) seems very promising for the study of muscle. It was shown that this technique provides an accurate estimation of individual muscle force in passive (i.e., stretching) and active (i.e., contractions) conditions. Recent studies that aim to study coordination during isometric contraction and changes in passive muscle tension using this technique will be reviewed. Perspectives for the study of dynamic contractions will also be briefly presented.

Abstract

Robotic therapy is a flagship example of the benefits of human-robot collaboration. However the 2010 American Heart Association guidelines for stroke care only endorsed robotic therapy for the upper extremity (UE), and not for the lower extremity (LE). In 2010, the US Veterans Administration similarly endorsed robotic therapy for UE but not for LE: “recommendation is made against routinely providing the [LE] intervention… At least fair evidence was found that the intervention is ineffective …” This apparent immaturity of LE robotic therapy reflects the fact that, to date, knowledge of human motor control has not been applied to LE robotic therapy. Knowledge of human motor control, sensing, and cognition has matured to the point that a fundamental theory of walking is now within reach. Walking can be composed of elementary actions, specifically submovements, oscillations, and mechanical impedance. This talk will discuss a proposal for a competent model of walking based on these elementary actions and code it into adaptive controllers that will allow multiple robotic devices to target human-like walking and rehabilitation.