Semi-Automated Analysis of Foveal Maturity in Premature and Full-Term Infants Using Handheld Swept-Source Optical Coherence Tomography

Abstract

Purpose: To develop a semi-automated method of measuring foveal maturity using investigational handheld swept source-optical coherence tomography (SS-OCT).

Methods: In this prospective, observational study, full-term newborns and preterm infants undergoing routine retinopathy of prematurity screening were imaged. Semi-automated analysis measured foveal angle and chorioretinal thicknesses at the central fovea and average two-sided parafovea by three-grader consensus, correlating with OCT features and demographics.

Results: One hundred ninety-four imaging sessions from 70 infants were included (47.8% girls, 37.6 ± 3.4 weeks postmenstrual age, 26 preterm infants with birth weight 1057 ± 325.0, gestational age 29.0 ± 3.0 weeks). Foveal angle (96.1 ± 22.0 degrees) steepened with increasing birth weight (P = 0.003), decreasing inner retinal layer thickness, and increasing gestational age, postmenstrual age, and foveal and parafoveal choroidal thickness (all P < 0.001). Inner retinal fovea/parafovea ratio (0.4 ± 0.2) correlated with increasing inner foveal layers, decreasing postmenstrual age, gestational age, and birth weight (all P < 0.001). Outer retinal F/P ratio (0.7 ± 0.2) correlated with ellipsoid zone presence (P < 0.001), increased gestational age (P = 0.002), and birth weight (P = 0.003). Foveal (447.8 ± 120.6 microns) and parafoveal (420.9 ± 109.2) choroidal thicknesses correlated with foveal ellipsoid zone presence (P = 0.007 and P = 0.01, respectively), postmenstrual age, birth weight, gestational age, and decreasing inner retinal layers (all P < 0.001).

Conclusions: Foveal development is dynamic and partially observed through semi-automated analysis of handheld SS-OCT imaging.

Translational relevance: Semi-automated analysis of SS-OCT images can identify measures of foveal maturity.

Figure 1.

Illustration of measurements by semi-automated analysis of foveal maturity. (A) Original OCT B-scan image encompassing the foveal center. (B) The program automatically segmented the ILM (green curved line), OPL (red curved line), RPE (blue line), and C/S junction (white curved line). Foveal angle was measured from the brim of the foveal depression (red dots on each side) to the center of the fovea (central red dot). Inner retina was measured from the ILM to the OPL. Outer retina was measured from the OPL to the RPE. Choroid was measured from RPE to the C/S junction. The program then aligned the B-scan with the RPE as the reference (i.e., making the RPE as a straight line as shown). The thickness measurements were performed at the three locations marked by central red dot and yellow asterisks for the foveal and parafoveal thickness, respectively. ILM, internal limiting membrane; OPL, outer plexiform layer; RPE, retinal pigment epithelium; C/S, choroid/sclera border. Scale bar = 1 mm.

Figure 2.

Semi-automated measurements of foveal maturity versus innermost layer and ellipsoid zone presence. (A) Mean F/P ratio at the inner retina strongly correlates with number of innermost layers present at the fovea. Mean F/P ratio at the outer retina strongly correlates with ellipsoid zone absence at the fovea. (B) Choroidal thickness at the fovea and parafovea correspond to absent inner retinal layers and present ellipsoid zone at the fovea. (C) Foveal angle increases with increasing innermost layers present at the fovea. See Figure 1 for an explanation of semi-automated measurements. Whiskers, standard deviation; F/P, fovea/parafovea ratio.

Figure 3.

Semi-automated measurements of foveal maturity versus postmenstrual age. (A) F/P ratio at the inner retina decreases with postmenstrual age (P < 0.001). (B) F/P ratio at the outer retina increases with postmenstrual age (P = 0.175). (C) Foveal angle decreases with postmenstrual age (P < 0.001). (D) Choroidal thickness at the fovea increases with postmenstrual age (P < 0.001). (E) Choroidal thickness at the parafovea increases with postmenstrual age (P < 0.001). F/P, fovea/parafovea ratio; w, weeks.

Figure 4.

Semi-automated measurements of foveal maturity versus gestational age. (A) Inner retinal F/P ratio decreases whereas outer retinal F/P ratio increases with gestational age (P < 0.001 and P = 0.002, respectively). (B) Foveal angle decreases with gestational age (P < 0.001). (C) Choroidal thickness at both the fovea and parafovea increase with gestational age (both P < 0.001). F/P, fovea/parafovea ratio; w, weeks.