RESUMO
PURPOSE: To demonstrate the demographic features, causative agents, and outcome of pediatric traumatic cataract surgery in eyes without posterior segment involvement at Assiut University Hospital, Upper Egypt. PATIENTS AND METHODS: This is a retrospective study on children (aged <18 years old) who underwent traumatic cataract surgery from January to June 2019. Children with posterior segment injury and those who did not complete 6 months of postoperative follow-up were excluded. The demographic features, mechanism and time of eye injury, clinical features, surgical approach, and outcome were recorded and analyzed. RESULTS: The study included 34 eyes of 34 children, 23 (68%) of them were boys. The mean age at the time of cataract surgery was 10±3.97 years. Twenty-one eyes sustained open globe injury (62%) with the most common cause of trauma was wooden sticks, while 13 eyes had closed globe injury (38%) with the most common cause of injury was thrown stones. The time interval between eye injury and cataract surgery ranged from 1 day to 9 years with a median of 2.05 months. Posterior chamber intraocular lenses were implanted in all eyes; in 33 eyes, the posterior chamber intraocular lenses were implanted primarily at the time of cataract extraction. Corrected distance visual acuity significantly improved from 2.63±0.66 LogMAR preoperatively, to 0.41±0.38 LogMAR postoperatively (p < 0.001). CONCLUSION: Pediatric traumatic cataract is commonly present in primary school age especially after open globe injury. Primary prevention through health awareness should target this age population. Useful vision can be regained with timely proper surgical intervention and posterior chamber intraocular lens implantation. CLINICALTRIALSGOV ID: NCT04630509.
RESUMO
Purpose: To assess the use of deep learning for high-performance image classification of color-coded corneal maps obtained using a Scheimpflug camera. Methods: We used a domain-specific convolutional neural network (CNN) to implement deep learning. CNN performance was assessed using standard metrics and detailed error analyses, including network activation maps. Results: The CNN classified four map-selectable display images with average accuracies of 0.983 and 0.958 for the training and test sets, respectively. Network activation maps revealed that the model was heavily influenced by clinically relevant spatial regions. Conclusions: Deep learning using color-coded Scheimpflug images achieved high diagnostic performance with regard to discriminating keratoconus, subclinical keratoconus, and normal corneal images at levels that may be useful in clinical practice when screening refractive surgery candidates. Translational Relevance: Deep learning can assist human graders in keratoconus detection in Scheimpflug camera color-coded corneal tomography maps.