The purpose of the study was to develop an evidence-based method for identification and classification of wheelchair sports (rugby) activities and performance during a match using fractal dimensions. The approach involves five wheelchair rugby athletes of different clas-sifications who were monitored during six different mat-ches using mobile devices with in-built accelerometers. The linear acceleration signals were processed using two different approaches based on calculating fractal dimen- sions. One was based on Rényi’s entropy, which produced the probability dimension (S 0 ), and the other was based on Hausdorff’s definition, resulting in the Hausdorff dimen-sion (D H ). When both dimensions were individually plotted as cumulative distribution plots, they offered two approa-ches to analyze the performance of a wheelchair rugby athlete. Combining the two dimensions produced a two-dimensional (2D) mapping that identified five different activities of each athlete during a wheelchair rugby match—(1) no activities, (2) low activities, (3) high-speed coasting, (4) high-speed pushing, and (5) extreme colli-sions. In the 2D mapping, four boundary lines separate the five different activities, which produced a template for each athlete. It was found in this research that the classification and skill level of the athlete had an effect on the boundary

EB00000002351K

Available