Evaluation of Optic Nerve Head Parameters, Retinal Nerve Fiber Layer, and Ganglion Cell Layer in Glaucoma using Deep Learning Classifier

ABSTRACT 

Aim: To develop a deep learning (DL) model for detection of glaucoma based on peripapillary retinal nerve fiber layer (pRNFL), ganglion cell layer (GCL), optic nerve head parameters using spectral domain optical coherence tomography (SD-OCT) and visual field parameters

Methods: 78 patient with glaucoma and 53 healthy subjects were recruited and split into training (%60) and test (%40) datasets. pRNFL, GCL, optic nerve head parameters and visual field parameters were used for the deep learning classifier. RapidMinerStudio9.2 was used for our deep learning model.

Results: In the test dataset, this deep learning system achieved an AUC of 0,817 with a sensitivity of % 96.

Conclusion: An SD-OCT and visual field based deep learning system can detect glaucomatous structural change with high sensitivity and specificity.

Key words: Glaucoma, deep learning, optical coherence tomography, artificial neural network