DRISHYA P. K. (STM21CS026)
FATHIMATHUL AIFA K. P. (STM21CS031)
SANDRA C. M. (STM21CS049)
SHEETAL MADHU (STM21CS057)

The increasing survival rates of premature infants have amplified the need for effective and accessible solutions to monitor and manage Retinopathy of Prematurity (ROP), a potentially blinding ocular condition. Traditional screening methods, often reliant on
specialized equipment and trained personnel, are complex and can lead to delays in diagnosis. This project introduces a web-based platform designed to streamline ROP detection and management through advanced deep-learning techniques. The system is
trained using a dataset with multiple input models which includes MobileNet, Inception, VGG, AlexNet, DenseNet, ResNet, and EfficientNet, to ensure accurate detection and staging of ROP. After training, the model that achieves the best accuracy is selected as the
final model, which is then used to process infant eye images uploaded by users, delivering immediate diagnostic feedback and detailed information on the condition’s implications.
To address challenges in rural areas, where healthcare infrastructure and awareness may be limited, the future implementation will integrate mobile phone- based ROP detection and telemedicine features. This approach aims to empower caregivers in remote
locations to conduct screenings and receive remote consultations, improving accessibility and ensuring timely intervention for premature infants at risk of ROP, regardless of their geographic location.