April 2014
Volume 55, Issue 13
Free
ARVO Annual Meeting Abstract  |   April 2014
Evaluation of a Portable Collision Warning Device for Visually Impaired Patients in an Indoor Obstacle Course
Author Affiliations & Notes
  • Shrinivas Pundlik
    Schepens Eye Research Institute, Boston, MA
  • Matteo Tomasi
    Schepens Eye Research Institute, Boston, MA
  • Gang Luo
    Schepens Eye Research Institute, Boston, MA
  • Footnotes
    Commercial Relationships Shrinivas Pundlik, Schepens Eye Research Institute, Mass Eye and Ear (P); Matteo Tomasi, None; Gang Luo, Schepens Eye Research Institute, Mass Eye and Ear (P)
  • Footnotes
    Support None
Investigative Ophthalmology & Visual Science April 2014, Vol.55, 2154. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Shrinivas Pundlik, Matteo Tomasi, Gang Luo; Evaluation of a Portable Collision Warning Device for Visually Impaired Patients in an Indoor Obstacle Course. Invest. Ophthalmol. Vis. Sci. 2014;55(13):2154.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose: Severe peripheral field loss leads to reduced mobility, increased frequency of collisions, and secondary injuries. We have developed a video camera based, pocket sized collision warning device that can detect impending collisions and issue audio warnings. Unlike a ranging sensor, the device processes videos to detect collisions by computing time-to-collision and collision point. An indoor high density obstacle course study is conducted to evaluate the utility of the device for collision avoidance.

Methods: 11 patients (8 with tunnel vision and 3 with hemianopia) walked in an obstacle course, which was set up in a form of 41-m long loop and consisted of 46 obstacles placed at various heights in a 20x6 m area. The subjects also encountered a pedestrian walking in opposite direction, whom they were instructed to yield to. Each subject was tested in two conditions, with and without the device. Number of collisions and preferred percentage of walking speed (PPWS) were compared within subjects. The subjects walked 4 loops continuously through the course in each condition. The order of the condition and the walking direction were counterbalanced for the tunnel vision group. Additionally, the subjects also performed a 1-back recall task while walking. A number was played via speakers every 3s, and as soon as the subjects heard one number they were asked repeat the last one. The subjects received training before the experiment in a 30 minute session about using the device and performing the 1-back recall task.

Results: When walking with the device, collisions were reduced for all the subjects on an average by 47% (p=0.0032). Comparing the two device conditions, no significant difference was observed in the walking speed (p=0.24) and secondary task performance (p=0.09). Mobility performance between the two conditions is not correlated with the secondary task error measures (R2=0.065). No significant learning effect was noticed when walking over multiple loops without the device (about 20% of total collisions recorded in the final loop).

Conclusions: Preliminary analysis indicates that the portable, video camera based collision warning device may help patients with peripheral field loss (especially tunnel vision) avoid collisions, while not significantly affecting their walking speed.

Keywords: 753 vision and action • 584 low vision • 549 image processing  
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×