Abstract
Abstract: :
Purpose: In Retinal Birefringence Scanning (RBS), a scanned annulus of polarized light is used to detect foveal fixation. Mathematical modeling of retinal birefringence scanning indicates that the foveal fixation signal originates from the retardance of radially symmetric nerve fibers surrounding the fovea. The GDx Nerve Fiber Analyzer is a polarized raster scan believed to map birefringence of the nerve fiber layer. Thus, both RBS and GDx assess retinal birefringence, and GDx images of the fovea produce a characteristic polarization cross. The purpose of this study was to determine how closely RBS parameters correlate with GDx findings. Methods: RBS was performed as previously described. Subjects were asked to fixate centrally and in at least four paracentral gaze positions. Pupil diameter and the intensity of reflected near-infrared light was measured with a frame grabber, and a pupil factor (pupil area x intensity) calculated. Phase shift of the RBS signal during central and paracentral fixation was measured. GDx images of the fovea were obtained by asking the subject to fixate on the center of the raster scan grid as the images were collected. The polarization cross was identified and its orientation measured. The intensity of the polarization cross was determined from the intensity profile of a 3° annulus. Results: Analysis was performed in 13 patients (26 eyes). The axis of the GDx polarization cross varied between eyes in all 13 subjects. While phase shift of the paracentral signal varied with gaze direction in all subjects, the absolute phase shift was not the same in different subjects. Correction of paracentral phase shift using the central phase shift parameter did not improve the predictability of phase shift except in 7 of 26 eyes. RBS phase shift during central fixation did not correlate with the orientation of the GDx polarization cross (r2 = -0.05). Signal strength during central fixation correlated mildly with the pupil factor (r = 0.322). RBS signal strength during central fixation correlated weakly with GDx polarization cross intensity (r = 0.27.) Conclusion: RBS parameters correlated only weakly with equivalent GDx measurements. Variations in corneal birefringence are likely to produce variability in the signals obtained using both instruments, but the fixed retarder used in GDx to compensate for corneal birefringence may interfere with correlation of the two techniques.
Keywords: 575 screening for ambylopia and strabismus • 406 eye movements • 408 eye movements: recording techniques