Figure 3 | Scientific Reports

Figure 3

From: Association of imaging factors derived from convolutional neural network with visual outcomes in age-related macular degeneration and polypoidal choroidal vasculopathy

Figure 3The alternative text for this image may have been generated using AI.

Representative ce of polypoidal choroidal vasculopathy (PCV) with pachychoroid. The patient is a 79-year-old woman with a baseline visual acuity of 20/40. For 31.0 months of follow-up, the eye was treated with intravitreal injections, including 14 aflibercept and 2 ranibizumab injections. (a) Baseline fundus photography shows subretinal fluid collection at macula. (b) Baseline indocyanine green angiography (ICGA) shows polypoidal lesion. (c) En face optical coherence tomography (En face OCT) scan at the final visit. Dilated choroidal vessels are prominent. (d) Enhanced depth imaging optical coherence (EDI OCT) at baseline. Choroidal layer is thickened with subfoveal choroidal thickness (SFCT) of 551 µm. (e) EDI OCT scan after 12 months from baseline. All areas of the lesions decreased, and visual acuity was maintained at 20/40. At the final visit, visual acuity improved to 20/32. (f) Upper panel shows predicted lesion corresponding baseline OCT B-scan by a convolutional neural network (CNN) (upper). Subretinal fluid (SRF, green color), subretinal hyperreflective material (SHRM, yellow color), and pigment epithelial detachment (PED, blue color) are noted. Each component could be quantified by converting the pixel counts into mm2, and the areas of SRF, SHRM, and PED are 0.650, 0.048, and 0.193 mm2, respectively. Lower panel shows predicted areas of SRF, SHRM and PED by the CNN as 0.091, 0.004, and 0.073 mm2, respectively.

Back to article page