Real-Time ML based Cataract Detection in Resource-Constrained Environment

Early detection and diagnosis of this disease are greatly aided by the initiative called cataract detection employing machine learning and image processing techniques. A cataract is a thick, hazy patch that forms in the lens of the eye. It gradually worsens until it obstructs eyesight. Due to deteriorating vision, this condition primarily affects the elderly, who eventually deal with a host of issues because of a lack of awareness, diagnosis, and detection. The input datasets are examined with the use of Convolutional Neural Networks and Support Vector Machine algorithms, which determine whether the input image of an eye has cataracts. An algorithm has been proposed to detect and predict the presence of cataracts in a person. Additionally, assess the trained models' performance using relevant measures like F1-score, recall, accuracy, and precision. Used strategies such as cross-validation to evaluate the models' capacity for generalization and reduce overfitting.