Deep Learning Models to Predict Primary Open-Angle Glaucoma Using Longitudinal Visual Field Measurements

Lei Liu

Abstract:

Glaucoma is a major cause of blindness and vision impairment worldwide and visual field (VF) tests are essential for monitoring the conversion of glaucoma. Existing research often uses VF data at a single time point to predict glaucoma; few explored the longitudinal trajectories. Additionally, many deep learning techniques treat the time-to-glaucoma prediction as a binary classification problem (glaucoma Yes/No), resulting in the misclassification of some censored subjects into the non-glaucoma category and decreased power. To tackle these challenges, we propose and apply several deep-learning approaches that naturally incorporate temporal and spatial information in longitudinal visual field data and predict time-to-glaucoma. The proposed methods’ prediction performance is validated on the large
Ocular Hypertension Treatment Study (OHTS) dataset. Extensive experiments show that the proposed LSTM and Bi-LSTM have better prediction performance than the traditional Cox proportional hazards model, ResNet50-LSTM, and CNN-LSTM methods.