Predicting refractive error from retinal fundus images using deep learning

Abstract

Purpose : Deep learning has show great promise in diagnosis and interpretation of retinal imaging with the potential to identify new image features to augment current medical practice. In this study we demonstrate this by training a model to estimate spherical equivalent refractive error from fundus photographs.

Methods : A deep convolutional neural network was trained to predict spherical equivalent refractive error from retinal fundus photographs from the UK Biobank (autorefraction) and AREDS (subjective refraction) clinical trial cohorts. Prediction accuracy was tested against documented refractive error measurements. We use soft-attention to identify image features that contributed most to model predictions.

Results : The resulting algorithm had a mean average error (MAE) of 0.56 diopters (95% CI: 0.55-0.56) for estimating spherical equivalent on the UK Biobank dataset and 0.91 diopters (95% CI: 0.89-0.92) for the AREDS dataset. The baseline expected MAE (obtained by predicting the mean of this population) is 1.81 diopters (95% CI: 1.79-1.84) for UK Biobank and 1.63 (95% CI: 1.60-1.67) for AREDS. Attention maps suggest that the foveal region is one of the most important areas that is used by the algorithm to make this prediction, though other regions also contribute to the prediction.

Conclusions : The ability to estimate refractive error with high accuracy from retinal fundus photos has not been previously known and demonstrates that deep learning can be applied to make novel predictions from medical images. In addition recent advances in obtaining retinal fundus photographs using mobile phones and inexpensive attachments suggest this work may be particularly relevant in regions of the world where autorefractors may not be readily available.

This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.