Development of an automatic evaluation system for balance assessment scales

The problem of a correct fall risk assessment is becoming more and more impelling with the increasing elderly population. However, despite this fact, no new tools have been developed to improve the ability to predict fall risk in individual patients and the clinical assessment procedures are based on subjective scoring of clinical balance scales. This work documents our current effort to develop a novel method to inspect balance control through a system for the automatic evaluation of exercises drawn from balance assessment scales. Our aim is to overcome classical limits found in these scales i.e. limited granularity and inter-/intra-rater reliability, to obtain objective scores and more detailed information allowing to predict fall risk.