Glaucoma Detection through Synergistic Combination of Image Processing and Deep CNN using MATLAB

This study introduces an innovative Glaucoma detection framework that significantly outperforms existing methodologies by employing a hybrid approach combining advanced image processing and Convolutional Neural Networks (CNN). The key to our approach is the precise segmentation of the optic cup using state-of-the-art image processing algorithms, which allows the CNN to accurately identify glaucomatous changes from retinal images. Our Glaucoma detection method had 95.45% sensitivity, 93.59% specificity, and 94.44% accuracy, surpassing other contemporary techniques such as the SVM Classifier, which showed an accuracy of 90%, and a Convolutional Neural Network approach by other researchers with an accuracy of 91.52%. These results affirm the potential of our method as a highly reliable tool for early Glaucoma diagnosis in clinical practice, promoting better patient outcomes through earlier treatment interventions. This study introduces an opportunity for enhancing glaucoma screening and early treatments through using of MATLAB software.