Researchers did this study to assess a deep learning classifier’s performance for GON differentiation from CON based on ganglion cell–inner plexiform layer (GCIPL) and RNFL spectral-domain optical coherence tomography (SD-OCT).
The researchers compiled Eighty SD-OCT image sets from 80 eyes of 80 patients with GON and 81 SD-OCT image sets from 54 eyes of 54 patients with CON for the study. The bottleneck features extracted from the GCIPL thickness map, GCIPL deviation map, RNFL thickness map, and RNFL deviation map were used as predictors for the deep learning classifier. The area under the receiver operating characteristic curve (AUC) was calculated to validate the diagnostic performance.
The deep learning system achieved an AUC of 0.990 with a sensitivity of 97.9% and a specificity of 92.6% in a fivefold cross-validation testing, which was significantly larger than the AUCs with the other parameters: 0.804 with temporal raphe sign, 0.815 with superonasal GCIPL, and 0.776 with superior GCIPL thicknesses.
The study concluded that the deep learning classifier is a better choice as it could outperform the conventional diagnostic parameters for GON discrimination and CON on SD-OCT.