Digital Enhancement of Neonatal Fundus Color Photography: Improving RoP Diagnosis through Image Processing

This study addresses the critical challenge of enhancing wide-angle contact color fundus photographs of premature neonates for improved Retinopathy of Prematurity (RoP) screening. RoP, a leading cause of preventable blindness in children, requires accurate and timely diagnosis for effective treatment. We investigated the efficacy of two image enhancement techniques: Contrast Limited Adaptive Histogram Equalization (ClaHE) and Automated Multi-Scale Retinex with Color Restoration (Automated MSRCR). These methods were applied to fundus images acquired by RetCam3 at Farabi Eye Hospital. The enhanced images were evaluated through qualitative assessment by expert ophthalmologists in RoP. Results demonstrated significant improvements in image quality, particularly for low-quality originals, with both CLAHE and Automated MSRCR increasing average image grades. A user interface (UI) was developed to facilitate the application of these enhancement techniques in clinical settings. This study highlights the potential of image enhancement in supporting more accurate and efficient RoP diagnosis. This research contributes to advancing neonatal eye care and potentially reducing RoP-related Blindness through improved screening processes.