A Wavelet-Based Methodology for Features Extraction in Postural Instability Analysis

摘要

The ageing population is posing new challenges regarding the health and social care systems. In the health domain, major threats are falls and postural instabilities. Focusing on postural instabilities, in this article, the Wavelet theory has been used for the sake of postural sway classification for its capability to highlight important postural control mechanisms in the human body. In particular, the proposed methodology is based on a 5-level Discrete Wavelet Transform and the use of features computed on the new transformed domain. Preliminary analysis have led to 100% S_e and 96% S_p in the classification task.