Abstract
Purpose:
To determine whether the patterns of fixational eye movements are altered when images are defocused (low pass-filtered).
Methods:
Six subjects (3 male, 3 female, 22 - 32 years of age, Caucasian origin) were asked to inspect an image on the computer screen which was either “in focus” or low-pass filtered by convolution with a point spread function that was equivalent to 5D of myopic defocus. The screen was at 79 cm distance and one pixel on the screen subtended a visual angle of 1.45 arcmin. The movements of the right eye were tracked at 166 Hz sampling rate in a highly magnified video image (39.9 pixels/mm) by custom-developed software (Visual C++ 8.0). A commercial IR camera (DMK 22AUC03, www.theimagingsource.com) was used and a single IR LED that generated a bright first Purkinje image. The angular resolution in natural eyes was determined to be 1-2 arcmin. Differences in eye movements variables were determined in the pooled data of all subjects using t-tests, analyzing the horizontal and vertical components of the eye movements separately.
Results:
The frequency of horizontal microsaccades (defined as saccades with an amplitude between 20 and 120 arcmin) increased from 3.3±2.9 to 6.3±4.4 in low-pass filtered images recorded over 4 secs (p<0.05). The frequency of vertical microsaccades increased with the low pass-filtered image from 1.1±1.7 to 3.1±3.6 in 4 secs (p<0.05). Vertical saccades were generally smaller than horizontal ones (p<0.01). The amplitudes of horizontal microsaccades increased from 48.1±17.5 arcmin to 68.4±29.4 arcmin (p<0.01). Amplitudes of vertical microsaccades did not change (34.8±13.4 to 39.8±16.8 arcmin). The horizontal “diameter of the distributions of gaze during fixation” (described by Cherici et al, J Vis 2012, who found about 18 arcmin) changed from 15.3±6.0 arcmin to 17.6±13.1 arcmin (p=0.34; one subject p<0.01, one subject p<0.05) and vertically from 11.7±5.1 to 12.3±6.8 arcmin (p=0.69; one subject p<0.01).
Conclusions:
The spatial frequency content of an image has impact on the pattern of fixational eye movements in humans. Accordingly, we expect that uncorrected myopic subjects have different patterns of fixational eye movements.
Keywords: 522 eye movements •
719 scene perception •
717 space and scene perception