RESUMEN
In light of growing interest in using emerging large language models (LLMs) for self-diagnosis, we systematically assessed the performance of ChatGPT-3.5, ChatGPT-4.0, and Google Bard in delivering proficient responses to 37 common inquiries regarding ocular symptoms. Responses were masked, randomly shuffled, and then graded by three consultant-level ophthalmologists for accuracy (poor, borderline, good) and comprehensiveness. Additionally, we evaluated the self-awareness capabilities (ability to self-check and self-correct) of the LLM-Chatbots. 89.2% of ChatGPT-4.0 responses were 'good'-rated, outperforming ChatGPT-3.5 (59.5%) and Google Bard (40.5%) significantly (all p < 0.001). All three LLM-Chatbots showed optimal mean comprehensiveness scores as well (ranging from 4.6 to 4.7 out of 5). However, they exhibited subpar to moderate self-awareness capabilities. Our study underscores the potential of ChatGPT-4.0 in delivering accurate and comprehensive responses to ocular symptom inquiries. Future rigorous validation of their performance is crucial to ensure their reliability and appropriateness for actual clinical use.
RESUMEN
(1) Background: Intravitreal anti-vascular endothelial growth factor (anti-VEGF) is an established treatment for center-involving diabetic macular edema (ci-DME). However, the clinical response is heterogeneous. This study investigated miRNAs as a biomarker to predict treatment response to anti-VEGF in DME. (2) Methods: Tear fluid, aqueous, and blood were collected from patients with treatment-naïve DME for miRNA expression profiling with quantitative polymerase chain reaction. Differentially expressed miRNAs between good and poor responders were identified from tear fluid. Bioinformatics analysis with the miEAA tool, miRTarBase Annotations, Gene Ontology categories, KEGG, and miRWalk pathways identified interactions between enriched miRNAs and biological pathways. (3) Results: Of 24 participants, 28 eyes received bevacizumab (15 eyes) or aflibercept (13 eyes). Tear fluid had the most detectable miRNA species (N = 315), followed by serum (N = 309), then aqueous humor (N = 134). MiRNAs that correlated with change in macular thickness were miR-214-3p, miR-320d, and hsa-miR-874-3p in good responders; and miR-98-5p, miR-196b-5p, and miR-454-3p in poor responders. VEGF-related pathways and the angiogenin-PRI complex were enriched in good responders, while transforming growth factor-ß and insulin-like growth factor pathways were enriched in poor responders. (4) Conclusions: We reported a panel of novel miRNAs that provide insight into biological pathways in DME. Validation in larger independent cohorts is needed to determine the predictive performance of these miRNA candidate biomarkers.
RESUMEN
AIM: To present a novel approach to treatment of dural carotid-cavernous fistulas via the medial ophthalmic vein. DESIGN: Retrospective case series. MATERIAL AND METHODS: In this retrospective case series, we present 2 patients (3 eyes) with Type C dural CCFs, who had failed cannulation via the conventional transfemoral route and the transorbital superior ophthalmic vein approach. They subsequently underwent CCF occlusion via an anterior orbital approach through the medial ophthalmic veins, at the Department of Ophthalmology, National University Hospital Singapore. CCF occlusion was confirmed intraoperatively using angiography. Both patients were evaluated postoperatively for best-corrected visual acuity and resolution of clinical signs and symptoms. RESULTS: Successful occlusion of CCFs via the medial ophthalmic veins were achieved in all three orbits, with excellent visual and cosmetic outcomes postoperatively. CONCLUSION: Dural CCFs may potentially lead to severe visual dysfunction and should be diagnosed and treated promptly. When all venous routes have been exhausted, the transorbital approach via the medial ophthalmic vein remains an excellent and viable alternative to access the fistula. Close cooperation between the orbital, anesthetic and radiological teams is essential in ensuring success of the operation.
