Keratoviz-A multistage keratoconus severity analysis and visualization using deep learning and class activated maps

<span lang="EN-US">The detection of keratoconus has been a difficult and arduous process over the years for ophthalmologists who have devised traditional approaches diagnosis including slit-lamp examination observation thinning corneal. The main contribution this paper is using deep learning models namely Resnet50 EfficientNet to not just detect whether an eye infected with or but also accurately stages infection mild, moderate, advanced. dataset used consists corneal topographic maps pentacam images. Individually achieved 97% 94% accuracy on dataset. We employed class activated (CAM) observe help visualize which areas images are utilized when making classifications different keratoconus. Using predict severity can drastically speed up provide accurate results at same time.</span>