Possibilities for real-time DFA based injury detection and skill level differentiation.

摘要

Detrended fluctuation analysis (DFA) is applied to running stride time series for the detection of long range correlations. However, there is no concurrent real-time data collection and DFA feedback mechanism. We aimed to verify a running analysis system for production of real-time DFA a values. Data were collected utilising an accelerometer, attached to the tibia. The accelerometer data were transmitted to MATLAB for processing. Results demonstrated DFA a value output < 2 seconds post individual running Epochs, and < 5 seconds post run completion. Our running analysis system provides rapid advanced variability information, important in both training, and injury prevention.