The Fixational Stability of the Human Eye Measured by Imaging the Cone Mosaic

Abstract

Abstract: : Purpose: Most previous methods that have measured the fixational stability of the human eye have done so indirectly by monitoring the movement of the front of the eye. We have developed an extremely accurate technique which monitors the movement of the cone mosaic directly, as it is ultimately the structure that determines fixation stability. Methods: The cone mosaics of six subjects who were attempting to fixate accurately were imaged with the Rochester Adaptive Optics Ophthalmoscope. Images were acquired at an eccentricity of 1.25 degrees from the foveal center. The fixation target was a Maltese cross and the subject's head was stabilized with a bite bar. We avoided motion of the retinal image of the fixation target with respect to the camera, such as might have been caused by small head movements, by putting the fixation target and the retinal imaging camera in conjugate planes, using a common artificial pupil, and using the same wavelength (550 nm). The subject initiated the adaptive optics correction procedure, which took 0.25 to 0.5 sec, and was immediately followed by image acquisition with a flash duration of 4ms. Results were similar in experiments in which the subject directly initiated the flash. The images were interpolated by a factor of 4 and cross-correlated to determine the translation of the retina from image to image. The image shift was determined by picking the peak pixel of the cross-correlation function. The image shift was then converted to a measure of arc minutes and the standard deviation of retinal image position across at least 59 trials per subject was used to characterize fixation stability. Results: The standard deviation ranged from 1.5 to 3.8 minutes of arc in the horizontal and vertical directions with a mean across subjects of 2.2 minutes, in agreement with previous studies. Conclusions: Due to the optical arrangement and the use of adaptive optics, this method is much more accurate than all previous methods. The relative position of the retina was determined with an accuracy that was at least 10 times smaller than a single cone photoreceptor. These measurements determine the maximum accuracy with which a stimulus can be delivered to a specific location on a subject's retina without tracking the eye.

Keywords: physiological optics • eye movements: recording techniques • eye movements