Purpose: To validate the performance of a commercially available, CE-certified deep learning (DL) system, RetCAD v.1.3.0 (Thirona, Nijmegen, The Netherlands), for the joint automatic detection of diabetic retinopathy (DR) and agerelated macular degeneration (AMD) in colour fundus (CF) images on a dataset with mixed presence of eye diseases. Methods: Evaluation of joint detection of referable DR and AMD was performed on a DR-AMD dataset with 600 images acquired during routine clinical practice, containing referable and non-referable cases of both diseases. Each image was graded for DR and AMD by an experienced ophthalmologist to establish the reference standard (RS), and by four independent observers for comparison with human performance. Validation was furtherly assessed on Messidor (1200 images) for individual identification of referable DR, and the Age-Related Eye Disease Study (AREDS) dataset (133 821 images) for referable AMD, against the corresponding RS. Results: Regarding joint validation on the DR-AMD dataset, the system achieved an area under the ROC curve (AUC) of 95.1% for detection of referable DR (SE = 90.1%, SP = 90.6%). For referable AMD, the AUC was 94.9% (SE = 91.8%, SP = 87.5%). Average human performance for DR was SE = 61.5% and SP = 97.8%; for AMD, SE = 76.5% and SP = 96.1%. Regarding detection of referable DR in Messidor, AUC was 97.5% (SE = 92.0%, SP = 92.1%); for referable AMD in AREDS, AUC was 92.7% (SE = 85.8%, SP = 86.0%). Conclusion: The validated system performs comparably to human experts at simultaneous detection of DR and AMD. This shows that DL systems can facilitate access to joint screening of eye diseases and become a quick and reliable support for ophthalmological experts.

No abstract

Background: Retrograde transsynaptic degeneration (RTSD) of the retinal ganglion cells and retinal nerve fiber layer after postgeniculate injury has been well documented, but to the best of our knowledge, associated retinal microvascular changes have not been examined. The purpose of our study was to assess vessel density (VD) at macular and peripapillary regions in patients with RTSD. Methods: Cross-sectional study including 16 patients with homonymous visual field defects secondary to unilateral postgeniculate visual pathway injury and 18 age-matched controls. All participants were examined with AngioVue optical coherence tomography angiography to measure the peripapillary vessel density and macular vessel density (pVD/mVD) as well as the peripapillary retinal nerve fiber layer (pRNFL) and macular ganglion cell complex (GCC) thicknesses. The pRNFL and macular ganglion cell–inner plexiform layer (GCIPL) thicknesses also were evaluated using Cirrus OCT. A normalized asymmetry score (NAS) was calculated for GCIPL and GCC thickness, and mVD. Results: Average pRNFL and macular GCIPL/GCC thicknesses were significantly thinner in both eyes of patients compared with control eyes (all P ≤ 0.05). Eight patients (50%), who showed a RTSD of the GCIPL map, had a relative thinning of the GCIPL/GCC ipsilateral to the brain lesion in both eyes (represented by a positive GCIPL-NAS/GCC-NAS). The mean pVD and mVD also were significantly reduced in patients (all P ≤ 0.05). There was a strong correlation between GCIPL-NAS/GCC-NAS and mVD-NAS index in both eyes (all r > 0.7, P = 0.001). Furthermore, there was a similar spatial pattern of damage for the macular GCC thickness and VD values. Conclusions: We demonstrated a significant VD decrease in peripapillary and macular areas of patients with RTSD because of postgeniculate lesions. The structural and microvascular asymmetry indexes were significantly correlated. These findings provide new insights regarding transsynaptic degeneration of the visual system.