IDENTIFICATION OF POST-COVID EFFECT ON CHILDREN RETINA USING CLASSIFICATION
DOI:
https://doi.org/10.62643/ijerst.2026.v22.n1(2).pp296-300Keywords:
Artificial Intelligence; Medical Imaging; Post-COVID; Pediatric Healthcare; Deep Learning; Transfer Learning; Explainable AI; Clinical Decision SupportAbstract
Post-COVID complications in pediatric patients present significant healthcare challenges, particularly in retinal health assessment. This research presents an AI-powered clinical decision support system for identifying PostCOVID retinal effects in children using fundus image classification. The system employs ResNet50 transfer learning with a novel 3-phase progressive fine-tuning strategy on the APTOS 2019 dataset containing 3,662 high-resolution fundus images. The model achieved medicalgrade performance with AUC-ROC of 0.9824, accuracy of 92.5%, sensitivity of 89.9%, and specificity of 94.3%, exceeding clinical deployment thresholds. Grad-CAM explainability visualization enables transparent AI decision-making for clinical trust. The system includes a comprehensive web-based interface for real-time analysis and automated clinical report generation, demonstrating readiness for clinical validation and deployment
Downloads
Published
Issue
Section
License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
