Experimental Approach to Identify Retinal Diseases based on Local Binary Pattern Screening with Biomedical Image Assesment

Retinal disease detection is a crucial area in biomedical imaging and ophthalmology, aiming to diagnose conditions such as diabetic retinopathy, glaucoma, and cataracts at an early stage. This study presents an Experimental Approach to Identify Retinal Diseases based on Local Binary Pattern (LBP) Screening with Biomedical Image Assessment, integrating texture-based feature extraction, biomedical image segmentation, and hybrid machine learning models. The methodology involves preprocessing techniques such as contrast stretching and Gaussian blur for noise reduction, followed by LBP feature extraction and Recursive Feature Elimination (RFE) for optimal feature selection. A hybrid Artificial Neural Network-Support Vector Machine (ANN-SVM) model is implemented for classification. The proposed method demonstrates an accuracy of 93.5%, surpassing traditional models such as SVM (85.4%), Random Forest (88.9%), and ANN (90.2%). Comparative analysis of LBP-based and non-LBP-based techniques further validates the efficiency of texture-based screening. The results suggest that integrating LBP descriptors with hybrid machine learning significantly improves diagnostic precision. Future work can explore deep learning architectures for enhanced feature representation and real-world clinical deployment. The proposed model holds significant potential for automated retinal disease screening, improving early diagnosis and patient outcomes.