Lisbon 2017 Delegate Registration Programme Exhibition Virtual Exhibition Satellites OneWorld Travel Discount
escrs app advert

Determining the predictive ability of keratoconus screening indices of three Scheimpflug-based devices to distinguish keratoconus from normal corneas

Search Title by author or title

Session Details

Session Title: Imaging Anterior Segment

Session Date/Time: Sunday 08/10/2017 | 16:00-18:00

Paper Time: 17:10

Venue: Meeting Center Room I

First Author: : U.Choudhary INDIA

Co Author(s): :    L. Kaweri   R. Shetty                 

Abstract Details


To compare the predictive ability of three Scheimpflug devices in differentiating keratoconus and subclinical keratoconus from normal corneas.


Narayana Nethralaya Eye Institute 121/C Chord Road, Rajajinagar 1st R Block, Bangalore 560010, India.


This retrospective comparative study included 84 normal, 70 eyes with subclinical keratoconus, and 92 keratoconic eyes. In each eye, keratoconus-screening indices were evaluated using Pentacam Galilei and Sirius. Receiver operating characteristic (ROC) curves were calculated and quantified using the area under the curve (AUC).


Highest sensitivity was seen for average pachymetric progression index (RPIavg) and maximum Ambrosio relation thickness (ARTmax) (90.2% each) on Pentacam and surface asymmetry index (SAI) on Galilei (89.1%). Highest specificity was for index of height asymmetry(IHA) on pentacam (100%) and symmetry index back (Slb) and 8mm root mean square per unit area (RMS/A) back (100% each) on Sirius. To distinguish subclinical-keratoconus from normals, highest sensitivity was for keratoconus index (100%) on Pentacam and CSI on Galilei (98.6%), and the highest specificity was seen for IHA and Curvature radius (100%) on Pentacam, Slb and 8mm RMS/A back (100% each) on Sirius.


Keratoconus indices measured by all three machines can effectively differentiate keratoconus from normal corneas. However, a combination of different data is required to distinguish subclinical keratoconus.

Financial Disclosure:


Back to previous