•   
    AAO 2021
    03:37
    Cornea/External Disease, Diagnostic & Imaging , Endothelial Keratoplasty

    In this interview from AAO 2021, Dr. Pooja Khamar describes a new artificial intelligence (AI)-based smartphone application she and her colleagues developed to help identify patients at risk of keratoconus and suggest optimal management. Her team used machine learning to teach the AI algorithm to weigh options, including crosslinking and keratoplasty, based on a 5-point nomogram published previously. In testing with more than 3,500 patient cases, the app was able to identify disease progression and suggest an appropriate treatment with 96% sensitivity and specificity.