Automatic detection of keratoconus on Pentacam images using feature selection based on deep learning

Abstract Today, corneal refraction, height, and thickness data, which are required in the diagnosis of keratoconus, can be obtained with tomography devices. Pentacam four map display presenting this data is one most basic options keratoconus. In article, an artificial intelligence‐based method using images proposed to distinguish keratoconus from healthy eyes. Axial/sagittal curvature, back elevation, front a total 341 corneas İnönü University ophthalmology clinic as set were given input AlexNet, deep learning models, feature vectors each image combined. The effective features determination determined by applying ReliefF, minimum‐redundancy‐maximum‐relevance (mRMR) Laplacian algorithms, widely used extraction vector. These classified support vector machine (SVM) classifier, has high performance binary classification. accuracy, specificity, sensitivity detection found 98.53%, 99.01%, 98.06%, respectively. developed model clinician evaluate cornea detect difficult through subjective assessments, especially subclinical early stages disease.