Deep Learning Approaches Predict Glaucomatous Visual Field Damage from OCT Optic Nerve Head En Face Images and Retinal Nerve Fiber Layer Thickness Maps

Abstract

Purpose: To develop and evaluate a deep learning system for differentiating between eyes with and without glaucomatous visual field damage (GVFD) and predicting the severity of GFVD from spectral domain OCT (SD OCT) optic nerve head images.

Design: Evaluation of a diagnostic technology.

Participants: A total of 9765 visual field (VF) SD OCT pairs collected from 1194 participants with and without GVFD (1909 eyes).

Methods: Deep learning models were trained to use SD OCT retinal nerve fiber layer (RNFL) thickness maps, RNFL en face images, and confocal scanning laser ophthalmoscopy (CSLO) images to identify eyes with GVFD and predict quantitative VF mean deviation (MD), pattern standard deviation (PSD), and mean VF sectoral pattern deviation (PD) from SD OCT data.

Main outcome measures: Deep learning models were compared with mean RNFL thickness for identifying GVFD using area under the curve (AUC), sensitivity, and specificity. For predicting MD, PSD, and mean sectoral PD, models were evaluated using R² and mean absolute error (MAE).

Results: In the independent test dataset, the deep learning models based on RNFL en face images achieved an AUC of 0.88 for identifying eyes with GVFD and 0.82 for detecting mild GVFD significantly (P < 0.001) better than using mean RNFL thickness measurements (AUC = 0.82 and 0.73, respectively). Deep learning models outperformed standard RNFL thickness measurements in predicting all quantitative VF metrics. In predicting MD, deep learning models based on RNFL en face images achieved an R² of 0.70 and MAE of 2.5 decibels (dB) compared with 0.45 and 3.7 dB for RNFL thickness measurements. In predicting mean VF sectoral PD, deep learning models achieved high accuracy in the inferior nasal (R² = 0.60) and superior nasal (R² = 0.67) sectors, moderate accuracy in inferior (R² = 0.26) and superior (R² = 0.35) sectors, and lower accuracy in the central (R² = 0.15) and temporal (R² = 0.12) sectors.

Conclusions: Deep learning models had high accuracy in identifying eyes with GFVD and predicting the severity of functional loss from SD OCT images. Accurately predicting the severity of GFVD from SD OCT imaging can help clinicians more effectively individualize the frequency of VF testing to the individual patient.

Figure 1:

Examples images for an eye with glaucomatous visual field damage and an eye without this damage. A single B-scan from the ONH cube scan is shown with the SALSA RNFL segmentation illustrated (top). The RNFL thickness map, RNFL enface image, and CSLO image are also shown (bottom).

Figure 2:

An illustration of the VF sectors used in this analysis (left) and their corresponding mapping on the ONH (right). These sectors are taken from Garway-Heath et al.

Figure 3:

Receiver operating characteristic curves in identifying GVFD eyes. The deep learning model based on RNFL enface images achieved the highest AUC of 0.88, significantly (p < 0.05) higher than the any other model.

Figure 4:

Heat maps created by occlusion testing that highlight informative image regions are shown for deep learning models based on RNFL thickness maps (A), RNFL enface images (B), and CSLO images (C). Color intensity indicates the amount of contribution to model classification of GVFD, prediction of global VF metrics (MD, PSD), and sectoral VF PD (central, temporal, inferior, inferior nasal, superior, superior nasal.

Figure 5:

Example RNFL thickness maps, RNFL enface images, and CSLO images for which the deep learning models produces predictions resulting in a true positive (A), true negative (B), false positive (C), and false negative (D).