A Tool for Automated Fall Risk Classification

Werner Kurschl, Michael Neuhold

Research output: Chapter in Book/Report/Conference proceedingsConference contributionpeer-review


For elderly it is more likely to fall and a fall can lead to severe injuries. Furthermore there is an ongoing discussion, if manual methods without any sensing and computing hardware or automated methods with the support of sensors and computing hardware are better in classifying elderly in regard of their risk to fall. Beside of these discussions, new sensing hardware allows the improvement of existing approaches so that automated fall risk classification can i) use very cheap pressure-sensing hardware, ii) can easily be used by nursing staff, and iii) keep the assessment time short. We present in this paper a tool for automated fall risk classification, which applies the leaning test. We also show that our approach with low measuring frequency and the rather coarse resolution of surface under pressure is able to achieve repeatable and reliable classification results.

Original languageEnglish
Title of host publicationITNG2010 - 7th International Conference on Information Technology
Subtitle of host publicationNew Generations
PublisherIEEE Computer Society’s Conference Publishing Services
Number of pages7
ISBN (Print)9780769539843
Publication statusPublished - 2010
Event7th International Conference on Information Technology : New Generations - Las Vegas, United States
Duration: 12 Apr 201014 Apr 2010

Publication series

NameITNG2010 - 7th International Conference on Information Technology: New Generations


Conference7th International Conference on Information Technology : New Generations
Country/TerritoryUnited States
CityLas Vegas
Internet address


  • Automated method
  • Center of body mass
  • Center of pressure
  • Flat pressure sensor
  • Mobility
  • Plastic-coated films
  • Reliable classification
  • Risk-scoring system


Dive into the research topics of 'A Tool for Automated Fall Risk Classification'. Together they form a unique fingerprint.

Cite this