Contrast Enhancement of Fundus Images by Employing Modified PSO for Improving the Performance of Deep Learning Models

Computer-Aided diagnosis (CAD) is a widely used technique to detect and diagnose diseases like tumors, cancers, edemas, etc. Several critical retinal diabetic retinopathy (DR), hypertensive (HR), Macular degeneration, retinitis pigmentosa (RP) are mainly analyzed based on the observation of fundus images. The raw images inferior quality represent minor changes directly. To analyze in vasculature or apply advanced disease detection algorithms, image should be enhanced enough visibly present vessel touristy. performance deep learning models for diagnosing these highly dependent accurate segmentation Specifically, vessels segmentation, challenging due low contrast, varying widths, branching, crossing vessels. For contrast enhancement, various retinal-vessel methods image-contrast enhancement as pre-processing step, which can introduce noise an affect detection. Recently, numerous studies applied Contrast Limited Adaptive Histogram Equalization (CLAHE) but with default values contextual region clip limit. In this study, our aim improve both supervised unsupervised machine by applying modified particle swarm optimization (MPSO) CLAHE parameter tuning, specific focus optimizing limit regions. We subsequently assessed capabilities optimized version using standard evaluation metrics. achieved MPSO-based demonstrating its real impact model semantic results proved positive sensitivity models, important. By proposed approach publicly available databases {DRIVE STARE}, we sensitivity, specificity accuracy {0.8315 0.8433}, {0.9750 0.9760} {0.9620 0.9645}, respectively.