Constancy of retinal vascular bifurcation geometry across the normal fundus and between venous to arterial bifurcations

Purpose: : 1)To assess the variability of retinal vascular bifurcations values across an individual fundus and between fellow eyes in normal subjects 2) To assess whether measures of retinal venous bifurcation geometry are optimised in a similar pattern as arterial bifurcations in normal subjects. Methods: : Forty–five degree nasal and temporal digital fundal images of 30 normal subjects (age range: 25–53 years) were assessed (60 images in total). Approximately 20 clearly defined arterial and venous bifurcations within each image were identified. Measures of bifurcation geometry (bifurcation angles and junction exponent) were obtained using custom designed semi–automated software. The spatial location of each bifurcation was defined in relation to the optic disc and the horizontal raphe of the fundus. Values were calculated and compared in relation to their location. Values of venous and arterial bifurcations within the same concentric region from the optic disc were also compared. Results: : There was no statistical difference in the geometrical values between the nasal and temporal half of the fundus. (Mean nasal bifurcation angles =81degrees, SD= 11, versus mean temporal angles = 82.2degrees, SD16)(p=0.5). There was also no difference between proximal angles (mean=81.7 degrees, SD=18) verses distal measurements (mean=81.47, SD= 13) (p= 0.6). The spread of the junction exponent was also found to be similar across the fundus. There was no significant difference between arterial and venous bifurcation angles across the fundus (Mean arterial bifurcation angles =81.5 degrees [SD= 16] versus mean venous bifurcation angles = 82.05 degrees [SD= 14] (p=0.64). The mean arterial junction exponent was 3.1 compared to a mean venous junction exponent of 2.6 showing no significant difference (p =0.25). Conclusions: : Measures of retinal vascular bifurcations do not change significantly across the retinal vascular tree. This sets a base line for future analysis of retinal vascular geometry in various vascular diseases that may affect differing regions of the retinal vasculature. Retinal venous bifurcations appear to follow the same constraints as retinal arterial bifurcations in normal subjects. Measurement of retinal venous bifurcation geometry in addition to arterial geometry may be valuable in assessment of the retinal vascular network.