Visual Control of Mobility in Low Vision

点击翻译按钮获取中文摘要

People with advanced vision impairment find walking in unfamiliar environments extremely difficult and, for this reason, often limit their independent travel 1, 2. This in turn increases their risk for loss of independence, depression, and an overall decline in health.3 Even people with moderate vision impairment have more bumps and stumbles and have a higher risk of falling than do persons with normal vision.2 While the effects of vision impairment on mobility performance have been identified, little is known about how vision impairments affect the mechanisms for the visual control of mobility. The objective of this research is to understand how vision impairment affects the ways in which information is explored and the source(s) of environmental information that is used to guide mobility. With this knowledge, more effective training protocols can be developed and smart mobility devices designed. The proposed research aims to determine how vision impairment affects visual exploration by analyzing travel time, gaze patterns and path trajectories of vision-impaired subjects as they walk to specified targets. In addition, a visual search paradigm4 will be employed to determine the role of low-level sensory factors on search time. The proposed research also aims to identify the cognitive factors that differentiate skilled from unskilled travelers with similar degrees of vision impairment. Two experiments will determine whether skilled travelers quickly learn contextual cues and/or direct their attention more to task- relevant information. Finally, the proposed research aims to determine the sources of visual information that travelers use to guide mobility and to determine if auditory information augments visual information to guide mobility for visually impaired travelers. Gaze patterns and path trajectories will be measured as subjects walk through routes where environmental information is manipulated (e.g. visual motion cue to distance). Several of the proposed experiments require a virtual environment (VE) to systematically control and/or manipulate environmental factors. Preliminary data demonstrate that VE is likely to be valid in studying exploratory behavior.