Automated macular segmentation can distinguish glaucomatous from compressive optic neuropathy

Abstract

Purpose To compare macular damage in glaucomatous optic neuropathy (GON) and compressive optic neuropathy (CON) and assess its diagnostic accuracy in distinguishing between diseases. Methods Observational, cross-sectional, single-center study. Patients with GON, CON, and healthy controls were included according to the eligibility criteria. An automated spectral-domain optical coherence tomography (SD-OCT) algorithm was used to segment the circumpapilary retinal nerve fiber layer (cpRNFL) and macula. The layer thickness was measured in each sector according to the Early Treatment Diabetic Retinopathy Study and the 6-sector Garway-Heath-based grids. Data was compared across all study groups, and the significance level was set at 0.05. Results Seventy-five eyes of 75 participants, 25 with GON, 25 with CON, and 25 healthy controls (CG), were included. Macular thickness was diminished in the ganglion cell complex of GON and CON patients compared to CG (p<0.05). The best Garway-Heath-based grid parameters for distinguishing GON and CON were the nasal-inferior (NI) and nasal-superior sectors and the NI/temporal inferior (TI) damage ratios in the macular ganglion cell (mGCL) and inner plexiform (IPL) layers. Moreover, the combination of the NI sector and NI/TI damage ratios in both layers had higher discriminative power (AUC 0.909; 95% CI 0.830–0.988; p<0.001) than combining parameters in each layer separately. Conclusion Our findings suggest that the evaluation of macular segmented layers damage by SD-OCT may be a helpful add-on tool in the differential diagnosis between GON and CON.

Optical Coherence Tomography (OCT): A Brief Look at the Uses and Technological Evolution of Ophthalmology

“…In simplified terms, the SD-OCT machine sends a beam of light through different tissue layers, e.g., the retina. The machine then uses advanced computer algorithms to process this information and create a highly detailed retina image [33].…”