Abstract
Abstract: :
Purpose: To investigate the distribution of retinal birefringence with eccentricity. To compare the distribution with known anatomical variation with eccentricity. Methods: Scanning laser polarimetry was performed with a GDx (LDT) in 20 normal subjects, 10 males and 10 females, age 17–55 yr. Linearly polarized light (780 nm) passes through a roughly 3 mm pupil, then is scanned pointwise to form a 15 X 15 deg raster focused on the retina, with 17.6 micron/pixel digital resolution. Differences among subjects in anterior segment polarization properties or illumination are nearly constant over the retinal image for a given subject, but vary across subjects. The radial amplitude variation over 20 input polarization angles at a detector with crossed polarization was fit with the sum of two sinewaves, one at the fundamental frequency of the macular cross (f) and the other at 1/2 f to allow for asymmetry and structures as the Raphe, with the phase of each sinewave variable for each eccentricity. Data were averaged for 3 neighboring rings. Analyses across subjects were performed by normalizing the eccentricity distribution by the amplitude of f at 3.21 deg. A second analysis compared two 1.5 deg samples, centered at 1.23 and 2.87 deg from the fovea. Results: The amplitude of sinusoidal modulation at f increased with increasing eccentricity until reaching a broad peak at 1.63 deg, on average, and was well–fit up to 1 deg with a straight line (p < 0.004) for 19 of 20 subjects. Amplitude at f was greater for 17 of 20 subjects at 1.41 deg from the fovea than at 3.21 deg, (F = 7.21, p= 0.0107). The difference in amplitude at these two eccentricities did not vary strongly with age for this normal sample of young to middle aged subjects (r=0.10), varying 0.09/decade. At the greater eccentricities, the amplitude at f typically decreased more gradually, but not necessarily linearly. Both the maximum amplitude of the fitted function and the amplitude at ½f were more variable with eccentricity, and not necessarily monotonic with peaks occurring in some subjects at eccentricities > 6 deg. When the amplitude at f was measured with a 1.5 deg sample, 16 of 20 subjects had greater amplitudes at 2.87 than at 1.23 deg, opposite to the result from the radial method. Conclusions: The birefringence that leads to the impression of a macular cross, thought to be due to light–tissue interactions of cornea and retina, extends well beyond the central macula where Henle fibers are radially oriented towards displaced ganglion cells. In the central 2 deg, normal subjects have well–quantified amplitude variations at the frequency of the macular cross.
Keywords: imaging/image analysis: non-clinical • macula/fovea • retina