An objective structural and functional reference standard in glaucoma

Abstract The current lack of consensus for diagnosing glaucoma makes it difficult to develop diagnostic tests derived from deep learning (DL) algorithms. In the present study, we propose an objective definition glaucomatous optic neuropathy (GON) using clearly defined parameters optical coherence tomography and standard automated perimetry. We then use proposed as reference a DL algorithm detect GON on fundus photos. A was trained photos, standard. performance evaluated independent test sample with sensitivity, specificity, area under receiver operating characteristic curve (AUC), likelihood ratios (LR). had 2118 photos 585 eyes 405 individuals. AUC discriminate between normal 0.92 sensitivity 77% at 95% specificity. LRs indicated that provided large changes in post-test probability disease majority eyes. evaluate high normal. newly used may increase comparability studies across devices populations.