Many sport disciplines require distinct whole-body movement actions and coordination for an effective and purposeful execution. The underlying mechanisms, which are often referred to as the technique, are difficult to describe and to approach scientifically. However, the interaction and reflection of one`s own technique is a crucial part for beginners to learn and for advanced to improve specific skills in their sport. This is particularly the case for alpine skiing, which unconventionally is dominated by propulsion through gravitational forces and a gliding ski-snow interface (LeMaster, 2010). Over the years, the developments in skiing technique and its elements (e. g. forward or backward leaning) have been captured in national curricula by experts demonstrating them visually or explaining them in written form (Österreichischer Skischulverband, 2015). To further access the elemental features, a translation into quantifiable measures might be beneficial. In the literature, this is often addressed by discrete kinematic variables (e. g. COM trajectories, joint angles or ground reaction forces), which are again difficult to interpret. Federolf et al. (2014) and Debertin et al. (2022) already showed the applicability of principal component analysis (PCA) for skiing technique evaluation, since it reveals principal movements, which can directly be associated with the elements described in the curricula. Besides, independent component analysis (ICA) has also been used to discriminate movement patterns (von Tscharner et al., 2013) and might be applicable for technique evaluations as well. The aim of the present study is to compare the capability of different algorithms and concepts to receive an optimized, interpretable and clearly separable set of variables in the shape of isolated principal (PC) and independent components (IC) to facilitate the access of technique elements in alpine skiing.
© Copyright 2023 dvs-Biomechanik 2023 Tagungsband. Julkaistu Tekijä Steinbeis-Edition. Kaikki oikeudet pidätetään. / Kääntänyt https://www.DeepL.com/Translator