Convolutional Neural Network Analysis for Glaucoma Detection in Retinal Image Analysis Compared with the K-Nearest Neighbors Algorithm’s Accuracy

The research involves rigorous training and testing procedures, leveraging α = 0.05, power = 85%, and N = 10 iterations through clincalc.com. With 216 retinal image samples, 80% are designated for training, while the remaining 20% are allocated for testing. This research evaluation comprises of 43 test sample images and 173 training sample images, and after processing samples in both algorithms, the obtained mean accuracy of novel CNN's (87.37%) which is superior over K-nearest neighbor (72.28%). Additionally, the research concludes with the relevance of using modern algorithms for boosting accuracy in medical imaging applications, with implications for better patient care and diagnostic skills in the field of ophthalmology.