Novel Pattern Reflection Topography Data Analyzed By Artificial Intelligence. Comparison Of Keratoconus Corneas To Normal, And Accuracy Vs. Scheimpflug Tomography.
Published 2023 - 41st Congress of the ESCRS
Reference: PO0212 | DOI: 10.82333/ma2y-cm88
Authors: Anastasios John John Kanellopoulos* 1 , Alexandros John Kanellopoulos 2 , Joshua A. Young 3 , Kathryn M. Hatch 4
1Ophthalmology,LaserVision Ambulatory Eye Surgery Unit,Athens,Greece;Ophthalmology,NYU Med School,New York,United States, 2Ophthalmology,LaserVision Ambulatory Eye Surgery Unit,Athens,Greece, 3Ophthalmology,New York University School of Medicine,New York,United States, 4Ophthalmology,Massachusetts Eye and Ear,Massachusetts,United States;Ophthalmology,Harvard Medical School,Massachusetts,United States
To evaluate the accuracy (specificity and sensitivity) of a novel corneal reflection topography device providing data to artificial intelligence for KCN diagnosis.
Laservision Ambulatory Surgical Center, Athens, Greece
120 individuals that underwent imaging by the novel topography device (Tilleron). The imaging data were processed by proprietary software to determine corneal normality vs. keratoconus. All cases were also imaged by Scheimpflug tomography (ST) imaging.
60 KCN cases and 60 non-KCN cases as defined by ST using the Amsler-Krumeich citeria were included. The Scheimplug tomography data rated the KCN cases in staged 1-3. The Tilleron data were 100 accurate in both KCN and normal cases with 100% sensitivity and 100% sensitivity
Artificial intelligence processing of this proprietary reflection topography pattern may provide accurate data in KCN diagnosis