Purchase this article with an account.
Joram Jacob van Rheede, Iain Wilson, Rose I Qian, Stuart Golodetz, Susan M Downes, Robert E MacLaren, Christopher Kennard, Stephen Lloyd Hicks; Mobility performance in low vision: Capturing the dynamics of target finding and obstacle avoidance across different tasks. Invest. Ophthalmol. Vis. Sci. 2014;55(13):4127.
Download citation file:
© 2017 Association for Research in Vision and Ophthalmology.
Recent developments in visual prosthesis and assistive technology research have increased the need for accurate visual mobility performance (VMP) measures in low vision (LV). Current approaches are based on the time participants take to follow a set course, overall walking speed, and the number of collisions. These measures do not generalise well across experimental setups, and do not capture the dynamics of VMP, which may provide useful information about the nature of visual constraints on mobility. We aim to address these issues by developing measures to dynamically assess VMP that generalise across different tasks, that do not require participants to follow a set course, and that are sensitive enough to differentiate between performance with and without assistive technology.
12 LV and 4 control participants performed target finding and obstacle avoidance tasks in a custom VMP assessment facility. We used cameras to track the position of participants in relation to obstacles and targets over time, and compared VMP with and without residual vision glasses (RVGs), a new assistive technology for LV.
We report that dynamic measures of walking behaviour provide more information about VMP than aggregate measures. Taking into account walking speed as a function of proximity to objects and targets and the heading direction of participants, it was possible to extract simple measures that reflect different aspects of VMP (see Figures). These measures could differentiate between LV and control participants, were able to quantify differences in VMP with and without RVGs, and did not require participants to follow a set course.
Current strategies for quantifying VMP can be improved upon by tracking the position of participants in relation to targets and obstacles over time to capture the dynamics of walking behaviour. This information can be used to generate simple output measures that can be generalised across tasks, and that are sensitive enough to differentiate between LV patients, monitor VMP over time, and investigate the potential benefit of assistive technology.
This PDF is available to Subscribers Only