Computer Vision System for the Blind Veteran
This is a feasibility study involving a small number of blind subject to evaluate the effectiveness of using a software and associated hardware components to assist the blind in identifying and locating objects of interests and to assist in way finding tasks. After the system has been successfully tested by the blindfolded engineers, it will be evaluated in a small number of blind subjects.
|Study Design:||Allocation: Non-Randomized
Endpoint Classification: Efficacy Study
Intervention Model: Crossover Assignment
Masking: Open Label
Primary Purpose: Supportive Care
|Official Title:||Computer Vision System for the Blind Veteran|
- time to complete a task (locate an object or to navigate to a door) [ Time Frame: effectiveness for locating objects will be determined after 2 years. effectiveness for navigation after 3 years. ] [ Designated as safety issue: No ]
|Study Start Date:||November 2011|
|Estimated Study Completion Date:||December 2014|
|Primary Completion Date:||December 2012 (Final data collection date for primary outcome measure)|
Experimental: Arm 1
the blind subject is asked to locate a randomly placed object with the assistance of the webcam/laptop computer.
Device: computer vision
a webcam-laptop based system to assist the blind in locating objects and in way finding.
No Intervention: Arm 2
The subject without the assist device can only locate the object by groping and random searching with hands.
This is primarily an engineering project to develop computer vision algorithms for an integrated system comprising of a laptop computer, webcam, gps, digital compass, microphone, and stereo earphones. The system will be voice controlled. The algorithm will be trained to identify and located objects and navigationally landmarks in real-time, previously selected by sighted developers. The subjects will be used to determine a list of objects to be identified and a list of useful information most important for way finding. After the system has passed objective engineering milestones such as the ability to correctly identify objects and landmarks with >99% accuracy and >90% sensitivity in <1 sec, subjects will be tested to determine whether the novel system will improve the ability of the blind to locate misplaced objects. Another test will be to determine whether the novel computer vision will decrease the time to safely navigate to the front door of a building.
Please refer to this study by its ClinicalTrials.gov identifier: NCT00920231
|United States, Georgia|
|Atlanta VA Medical and Rehab Center, Decatur|
|Decatur, Georgia, United States, 30033|
|United States, Maryland|
|Baltimore VA Medical Center VA Maryland Health Care System, Baltimore, MD|
|Baltimore, Maryland, United States, 21201|
|Principal Investigator:||Cha-Min Tang, PhD MD||VA Maryland Health Care System, Baltimore|