Fixation strategies revealed by the retinal imaging
23.446, Saturday, 17-May, 8:30 am - 12:30 pm, Banyan Breezeway
Girish Kumar1, Susana Chung1,2; 1School of Optometry, University of California, Berkeley, 2Vision Science Graduate Program, University of California Berkeley
It is well known that the resolving power of the retina is not uniform. Visual acuity is highest at the fovea and decreases at greater eccentricities. It is possible that normal subjects would use different locations on the retina to fixate at targets that have different resolution requirements, rather than always fixating using the retinal locus with the highest resolution. To test this hypothesis, we recorded the fixation behavior of 4 subjects using a Scanning Laser Ophthalmoscope (SLO) while simultaneously presenting fixation targets. This experimental setup allows us to determine the precise retinal location of the stimulus. We used two different types of stimuli: square dots that were 4, 8 and 16 arc minutes in size as well as letters that were 10, 20 and 40 arc minutes in height. Letters were presented singly or in groups of three (trigrams) or five (pentagrams) and were rendered using the Sloan font set. Subjects were instructed to fixate at the center of the target (square, single letter, trigram, pentagram) which was presented randomly within the 100 SLO imaging raster for 10 seconds. All sizes for each stimulus type were repeated 3 times. Videos of the subject's retina during each trial were recorded and eye movements were extracted from these videos using a cross-correlation technique. The fixation variability (measured using Bivariate Contour Ellipse Area (BCEA)) and retinal locus of fixation were calculated from the extracted fixation eye movements and compared across subjects. Fixation variability ranged from 0.04 to 0.76 deg2 across the different fixation conditions. Analysis of Variance did not reveal any significant statistical differences between conditions. Our results lead us to infer that for the range of stimulus sizes used in our study, normal subjects do not adopt different fixation strategies based on resolution requirements.