The Purpose of the study was to develop a deep residual learning algorithm to screen for glaucoma from fundus photography and measure its diagnostic performance compared to Residents in Ophthalmology. A training dataset consisted of 1,364 color fundus photographs with glaucomatous indications and 1,768 color fundus photographs without glaucomatous features. A testing dataset consisted of 60 eyes of 60 glaucoma patients and 50 eyes of 50 normal subjects. Using the training dataset, a deep learning algorithm known as Deep Residual Learning for Image Recognition (ResNet) was developed to discriminate glaucoma, and its diagnostic accuracy was validated in the testing dataset, using the area under the receiver operating characteristic curve (AROC). The Deep Residual Learning for Image Recognition was constructed using the training dataset and validated using the testing dataset. The presence of glaucoma in the testing dataset was also confirmed by three Residents in Ophthalmology. The deep learning algorithm achieved significantly higher diagnostic performance compared to Residents in Ophthalmology; with ResNet, the AROC from all testing data was 96.5 (95% confidence interval [CI]: 93.5 to 99.6)% while the AROCs obtained by the three Residents were between 72.6% and 91.2%.

No abstract

Background/aimsTo validate a deep learning algorithm to diagnose glaucoma from fundus photography obtained with a smartphone.MethodsA training dataset consisting of 1364 colour fundus photographs with glaucomatous indications and 1768 colour fundus photographs without glaucomatous features was obtained using an ordinary fundus camera. The testing dataset consisted of 73 eyes of 73 patients with glaucoma and 89 eyes of 89 normative subjects. In the testing dataset, fundus photographs were acquired using an ordinary fundus camera and a smartphone. A deep learning algorithm was developed to diagnose glaucoma using a training dataset. The trained neural network was evaluated by prediction result of the diagnostic of glaucoma or normal over the test datasets, using images from both an ordinary fundus camera and a smartphone. Diagnostic accuracy was assessed using the area under the receiver operating characteristic curve (AROC).ResultsThe AROC with a fundus camera was 98.9% and 84.2% with a smartphone. When validated only in eyes with advanced glaucoma (mean deviation value < −12 dB, N=26), the AROC with a fundus camera was 99.3% and 90.0% with a smartphone. There were significant differences between these AROC values using different cameras.ConclusionThe usefulness of a deep learning algorithm to automatically screen for glaucoma from smartphone-based fundus photographs was validated. The algorithm had a considerable high diagnostic ability, particularly in eyes with advanced glaucoma.