Keratoconus Progression Determined at the First Visit: A Deep Learning Approach With Fusion of Imaging and Numerical Clinical Data.

Purpose: Multiple clinical visits are necessary to determine progression of keratoconus before offering corneal cross-linking. The purpose of this study was to develop a neural network that can potentially predict progression during the initial visit using tomography images and other clinical risk factors. Methods: The neural network's development depended on data from 570 keratoconus eyes. During the initial visit, numerical risk factors and posterior elevation maps from Scheimpflug imaging were collected. Increase of steepest keratometry of 1 diopter during follow-up was used as the progression criterion. The data were partitioned into training, validation, and test sets. The first two were used for training, and the latter for performance statistics. The impact of individual risk factors and images was assessed using ablation studies and class activation maps. Results: The most accurate prediction of progression during the initial visit was obtained by using a combination of MobileNet and a multilayer perceptron with an accuracy of 0.83. Using numerical risk factors alone resulted in an accuracy of 0.82. The use of only images had an accuracy of 0.77. The most influential risk factors in the ablation study were age and posterior elevation. The greatest activation in the class activation maps was seen at the highest posterior elevation where there was significant deviation from the best fit sphere. Conclusions: The neural network has exhibited good performance in predicting potential future progression during the initial visit. Translational Relevance: The developed neural network could be of clinical significance for keratoconus patients by identifying individuals at risk of progression.

Retinal Toxicity after Initial Administration of Nivolumab and Ipilimumab.

Background: To present a rare case of a bilateral immune checkpoint inhibitor- (ICI-) induced photoreceptor injury with a bacillary layer detachment (BALAD) and a dissection of the photoreceptor inner and outer segment, accompanied by ICI-induced Vogt-Koyanagi-Harada- (VKH-) like uveitis after initial administration of nivolumab and ipilimumab. Case Presentation. A 52-year-old female with metastatic malignant cutaneous melanoma experiencing bilateral progressive visual acuity reduction, after treatment initiation with 1 mg/kg nivolumab and 3 mg/kg ipilimumab two weeks prior symptom onset. An extensive laboratory workup, including uveitis workup, onconeuronal and retinal antibodies, ruled out a paraneoplastic autoimmune disorder and a granulomatous disease. Furthermore, a B-scan was performed to exclude a posterior scleritis. Ensuing temporary treatment discontinuation of nivolumab and complete discontinuation of ipilimumab, treatment with high-dose systemic steroids was initiated, which resulted in alleviation of her symptoms and stability of ocular findings. Conclusions: ICIs can induce significant ocular side effects. As ocular inflammation can be well controlled using systemic steroids, treatment with ICIs can be continued whenever possible, in particular, if there is a good treatment response of the systemic malignancy.

[Photo-based examination for diabetic eye pathologies in a German ophthalmological practice without personal doctor-patient contact]. / Fotobasierte Untersuchung auf diabetische Augenveränderungen in einer deutschen Augenarztpraxis ohne direkten Arzt-Patienten-Kontakt.

BACKGROUND: An increasing number of patients suffering from diabetes require regular ophthalmological check-ups to diagnose and/or treat potential diabetic retinal disease. Some countries have already implemented systematic fundus assessments including artificial intelligence-based programs in order to detect sight-threatening retinopathy. The aim of this study was to improve the detection of diabetic fundus changes in Germany without examination by a doctor and to create an easy access to ophthalmological examinations. MATERIAL AND METHODS: In this prospective monocentric study 93 patients in need for a routine check-up for diabetic retinopathy were included. The study participants took up an offer of an examination (visual examination, non-mydriatic camera-based fundus examination) without doctor-patient contact. Patient satisfaction with the organization and examinations was assessed using a questionnaire. RESULTS: The mean age was 53.5 years (SD 13.6 years, 49.5% female) and 17 eyes (18.3%) showed a diabetic retinopathy which was detected using a camera-based examination. Within the small sample, no patient had to repeat the examination due to poor image quality. All categories of the questionnaire showed a good to very good satisfaction, indicating a high acceptance of the other examination form that took place at the ophthalmologist's premises. CONCLUSION: In our study in an ophthalmological practice a high level of acceptance among the patients interested in the screening for diabetic retinopathy without any direct patient-doctor contact was achieved. Our study shows a very good acceptance and feasibility. Future use of artificial intelligence in clinical practice may help to be able to screen many more patients as in this study imaging quality was very good.