Background and objectives: Spectral Domain Optical Coherence Tomography (SD-OCT) is a volumetric imaging technique that allows measuring patterns between layers such as small amounts of fluid. Since 2012, automatic medical image analysis performance has steadily increased through the use of deep learning models that automatically learn relevant features for specific tasks, instead of designing visual features manually. Nevertheless, providing insights and interpretation of the predictions made by the model is still a challenge. This paper describes a deep learning model able to detect medically interpretable information in relevant images from a volume to classify diabetes-related retinal diseases.Methods: This article presents a new deep learning model, OCT-NET, which is a customized convolutional neural network for processing scans extracted from optical coherence tomography volumes. OCT-NET is applied to the classification of three conditions seen in SD-OCT volumes. Additionally, the proposed model includes a feedback stage that highlights the areas of the scans to support the interpretation of the results. This information is potentially useful for a medical specialist while assessing the prediction produced by the model. Results: The proposed model was tested on the public SERI-CUHK and A2A SD-OCT data sets containing healthy, diabetic retinopathy, diabetic macular edema and age-related macular degeneration. The experimental evaluation shows that the proposed method outperforms conventional convolutional deep learning models from the state of the art reported on the SERI+CUHK and A2A SD-OCT data sets with a precision of 93% and an area under the ROC curve (AUC) of 0.99 respectively. Conclusions:The proposed method is able to classify the three studied retinal diseases with high accuracy. One advantage of the method is its ability to produce interpretable clinical information in the form of highlighting the regions of the image that most contribute to the classifier decision.

PurposeTo identify the factors predicting the visual and anatomical outcomes in eyes with central serous chorioretinopathy (CSCR) through 12 months.MethodsPatients with diagnosis of CSCR, either acute or chronic, were included in this multicentric, retrospective study. Demographic factors; systemic risk factors; central macular thickness (CMT), subfoveal choroidal thickness (SFCT), linear extent of ellipsoid zone (EZ) and interdigitation zone damage on optical coherence tomography; details of leak on fluorescein angiography and indocyanine green angiography were included as predictors of anatomical and visual outcomes. Regression analysis was performed to correlate the changes in best corrected visual acuity (BCVA) and resolution of disease activity.ResultsA total of 231 eyes of 201 patients with a mean age (49.7±11.8 years) were analysed. A total of 97 and 134 eyes were classified as acute and chronic CSCR. BCVA (0.35±0.31 to 0.24±0.34; p<0.001), baseline optical coherence tomography (OCT) parameters including CMT (p<0.001), subretinal fluid (SRF) height (p<0.001) and SFCT (p=0.05) showed a significant change through 12 months. Multivariate regression analysis showed change in CMT (p≤0.01) and SRF height at baseline (p=0.05) as factors predictive of good visual outcome. Logistic regression analysis revealed changes in both CMT (p=0.009) and SFCT (p=0.01) through 12 months to correlate with the resolution of disease.ConclusionOCT parameters such as changes in both CMT and SFCT along with subfoveal EZ damage can be predictive of disease resolution whereas changes in CMT and baseline SRF height correlate well with changes in BCVA through 12 months.

Purpose: To describe the clinical features of uveitis in children treated at two ophthalmological centers in Bogotá, Colombia, in a 13 year-period. Methods: Retrospective observational clinical record review of pediatric children with diagnosis of uveitis. Results: 310 children were evaluated, 51.9% were female, mean age of 10.1 years. Posterior uveitis was the most common location (58.7%), of insidious onset (87.4%) and chronic course (78.1%). The most common etiology was infectious (58.4%) caused by toxoplasmosis (76.8%). There was a statistically significant difference in visual acuity between anterior (20/68) and intermediate uveitis (20/70), compared to posterior uveitis (20/434) (p <0,05).Conclusions: This is the first study to report the clinical features of pediatric uveitis in Colombia, where infectious etiologies are the leading cause. It will improve awareness and knowledge of pediatric uveitis in developing countries, and contribute to the development of public health policies of pediatric visual health.

No abstract