Information Visualisation of Optimised Underhand Throw for Cybernetic Training

Abstract

We have previously proposed a training system that helps inexperienced athletes acquire skills through a repeated comparison of their electromyography (EMG) signals measured in real time to signals produced by an optimisation calculation using a musculoskeletal model. We refer to this training system as Cybernetic Training, which references a feedback-based signal produced artificially by optimisation of the model calculations. However, when considering many muscles, it is difficult to compare EMG signals and optimised signals because of the significant amount of information. The aim of this study was to develop a method to integrate significant amounts of human motion information to facilitate convenient perception during the Cybernetic Training. In the proposed method, a self-organising map (SOM) is employed to visualise the integrated motion data. Examples of visualisations include the motion data of an optimised underhand throw and that of human subjects wearing inertial and EMG sensors. We compared the optimised motion data and measured motion data using the obtained SOM.