Toward Low-Cost Balance Assessments for the Elderly

Sonya Allin1 , Cheryl Beach2 , Andrew Mitz3 , and Alex Mihailidis1
(1) The University of Toronto, Toronto, ON (2) Vancouver Island Health Authority, Victoria, B.C. (3) National Institutes of Health, Washington, D.C.


We present a feasibility study to determine the potential to cheaply automate items of the Berg Balance Scale (BBS), a popular tool for assessing functional balance, using low-fidelity motion data from an un-calibrated video camera. Berg Assessments of five elderly individuals were recorded in a community center setting both before and after participation in a falls prevention program. We demonstrate that a simple head tracker can generate movement statistics that distinguish levels of balance impairment in this subject pool.


Older adults, rehabilitation, balance assessment, Berg Balance Scale, computer vision


This study was funded by the Victoria Foundation and by Toronto Rehabilitation Institute.

Author Contact Information:

Sonya Allin, Ph.D., AT The University of Toronto, EMAIL: