Visual exploration with PCA dimension reduction to support fall-risk assessment of community-dwelling elderly

Purpose Identifying community dwelling elderly with high fall risk and sending them to hospitals tor further evaluation is important to keep and maintain older adults' quality ol lite. Comprehensive fall risk assessment is costly in terms of human labor and equipment. An alternative and more feasible approach is to conduct preliminary fall risk assessment through community services to identify elderly with elevated Kill risk and ask them to go to hospital for further evaluation. This paper proposes ,i visual exploration approach together with principle-component analysis (PCA) based dimension reduction to help medical professionals evaluate fall risk based on multi-dimensional data collected from preliminary fall risk evaluation tools. Method A total of 392 community-dwelling elderly from central Taiwan were evaluated from April 2014 to May 2015. After excluding participants aged