Abstract
Introduction: Individuals with macular degeneration are often unaware of information they miss in their central vision, likely due to perceptual filling in. In theory, making saccades into the scotoma can compensate for this. To determine the feasibility of this approach, we first tested whether normally sighted controls could direct saccades into an artificial scotoma. Method: Four observers made a same/different judgment on pairs of stimuli drawn from 24 selected silhouettes in the Snodgrass & Vanderwart image set (1980) and presented on opposite sides 8.5 degrees from fixation. The visible stimulus appeared gradually over 250 msec to avoid abrupt transients associated with the anti-saccade task. A peripheral artificial scotoma hid the other stimulus until a saccade was made towards it. Observers had up to 2 s to make a saccade, but the display was extinguished 300 msec after the first saccade. In Experiment 1 stimuli within a block were on the same axis spanning fixation, along either a cardinal or oblique axis. In experiment 2, cardinal and oblique axes were intermixed. Axis angles ranged between 0 and 360 degrees, in 45 degrees increments. Results: Three observers successfully directed their gaze into the scotoma and uncovered the hidden stimulus. The fourth observer did so on slightly more than half of the trials. Saccade latencies were longer when saccades targeted the scotoma, compared to when they targeted the visible stimulus, particularly when target axes were interleaved. There was no significant latency difference between targets on cardinal and oblique axes. Conclusion: Our results demonstrate that normally sighted observers can look into the blank region of an artificial scotoma, even when a distracting target is present. This oculomotor strategy is similar to an anti-saccade and requires a delayed and more deliberate saccade plan, especially when the location of the missing information is unknown in advance.
Meeting abstract presented at VSS 2014