Detection of Cataract, Glaucoma and Diabetic Retinopathy Eye Disease using Multiple Models

Preventing vision loss requires early identification of diabetic retinopathy, cataract, and glaucoma. This work classifies these disorders from retinal fundus pictures using a multimodel machine learning approach. CNN, Logistic Regression, Decision Trees, SVM, Random Forest, and Linear Regression models were trained on a dataset of 4,000 preprocessed pictures. Accuracy, precision, recall, and F1-score were used to assess performance; CNN’s robust feature-learning capacity allowed it to achieve the highest accuracy of 96.1%. The goal of this approach is to improve patient outcomes by facilitating remote and early diagnosis. The dataset will be enlarged in subsequent research, and explainable AI for clinical integration will be investigated.