Università degli Studi di Padova

Post-Doc, Dipartimento di Ingegneria dell'Informazione (DEI)

Thesis Title: Principal Component Analysis for motor skills characterization and individual monitoring in Sports Science

About

Sports biomechanics is the science field that applies the laws of mechanics and physics to athlete performance, in order to gain a greater understanding of motor skills through measurement, modelling and simulation. It aims to meet the growing demands within coaches, physicians and athletes, to quantitatively assess the performance motor characteristics. While in clinical “gait analysis”, standard protocols have been widely validated, in sports field the great amount of disciplines and the difficulty in standardizing movements have acted as a brake on the usage of very useful motion capture and analysis techniques which would allow quantitative analysis. This technique, even if offering great potential, is still in its infancy; substantial issues remain to be addressed, such as the need for subject-specific models and the tolerances within which changes in technique will not affect performance. The main task for any level of coaching is to construct a training program that will ensure continuous progression of an athlete whilst avoiding injuries. Therefore, there is the need for extensive indications that could let biomechanists exploit the potentials of motion analysis technologies by setting proper experimental protocols, by using effective data processing and analysis and, finally, by producing reports that may turn useful “on the field”.
The aim of sports research should be the identification of the peculiar characteristics and of the most proficient strategy for each athlete. There is the need to shift from a horizontal to a longitudinal experimental design, where the athlete is no more related to a normality range, but is compared with himself in different times during the training season. Since every athlete is characterised by his own abilities and deficiencies, trainers and coaches might get more effective results by applying and monitoring individual training programs rather than using the same strategy for every athlete.
Hence, the aim of this research was to find a way to get a robust representation of the athlete’s motor skills. Race walking has been chosen as the mean of investigation. It is a motor task that presents peculiar biomechanical and coordinative demands, particularly interesting for this study.
The purpose was to give a robust and complete characterization of the single athlete’s performance strategy. The comprehension and consequent depiction of individual motor behaviours is strongly affected by the presence of motor biovariability, which originates from many sources, and which is inherently present both intra and inter subjects. It does not allow an immediate comprehension of the individual motor status: it may hide the little progresses that the athlete achieves. Therefore standard data analysis techniques (mean, standard deviation, etc.) fail in extracting significative information from a large amount of kinematic and kinetic data.
Principal Component Analysis (PCA) is a multivariate statistical technique extremely effective in the study of human motion. It has been usually adopted to identify groups of inter-related variables, different walking patterns or gait forms, or to evaluate changes due to pathology, recovery or intervention. In this research three ways to apply PCA were compared: traditional PCA (t-PCA), functional PCA (f-PCA) and two-stage PCA (2-PCA). All the three methods allowed to exploit the correlation between multiple measures in the analysis of race walking data, thus allowing an overall characterization of race walking and providing important insight into each athlete peculiar characteristics. Extracting important information from motion data is a huge problem faced in gait analysis. Results are commonly interpreted subjectively from a large number of highly correlated, time-varying and constant variables. The techniques that were used in this study to detect biomechanical differences among athletes and to gain insight into the single athlete performance repeatability represented an objective method of simultaneously reducing and analysing many interrelated time varying measures.
The final purpose is trying to show how a complex and theoretical mathematical approach (PCA) can find practical application by matching results with field needs (motor skills characterisation) and by making information intelligible for practitioners. Therefore it may be the first step to help trainers in conceiving specific training sessions, taking into consideration peculiar athlete’s abilities and faults.