Abstract
Purpose: Recent data challenges the assumption that observers use optic flow at walking speeds in real environments (Rushton et al, 1998, Current Biology 1191). We compared mobility and eye movement behaviour in low vision observers with central (age-related macular disease) or peripheral (glaucoma) field loss. In the former, wide-field motion (optic flow) cues are spared, but central high resolution acuity is impoverished, in the latter, the opposite is true. This is a preliminary study investigating 2 normal observers and 5 patients.
Methods: Mobility, fixation behaviour and direction of gaze were measured with a video-based mobile head-mounted eye-tracker (ASL 501) in patients and age-matched controls while walking a 132m course. The course was divided into six sections, including corridors, stairs and road-crossings. Instructions were given at the beginning of each section. Gaze position was classified into scene categories that were analysed manually and computationally by cross-correlation of the image at fixation over time.
Results: Patients with central and peripheral field loss showed impaired mobility and fixation behaviour compared to age-matched controls. Performance was not dependent on the location of the visual field loss. Age-matched controls primarily fixated the pre-determined goal points. Patients had larger and more variable fixations and saccades, viewed more scene categories and took longer to reach the goal. This was confirmed with the objective cross-correlation analysis.
Conclusions: Loss of visual field impairs mobility in patients when walking an unfamiliar real environment, independent of the location of the scotoma. Age-matched controls directed their fixations mainly towards the goals, while patients fixated many other objects in the scene indicating that they spent longer searching for the goal or potential obstacles. Both high resolution central vision and wide field sensitivity are critical components of walking behaviour in real environments.