Time to reinvent the wheel: Bowman's surface topography of keratoconus eyes for improved detection

Session Details

Session Title: Anterior Segment Imaging I

Session Date/Time: Monday 24/09/2018 | 16:30-18:00

Paper Time: 17:27

Venue: Room A4

First Author: : R.Chandapura INDIA

Co Author(s): :    M. Salomao   P. Khamar   M. Francis   R. Ambrosio   R. Shetty   A. Sinha Roy     

Abstract Details

Purpose:

To evaluate the sensitivity (Se) and specificity (Sp) of anterior (A) and bowman's (B) surface topography parameters in normal, suspect and keratoconus (KC) eyes.

Setting:

Narayana Nethralaya Eye Hospital, Bangalore, India

Methods:

Anterior segment optical coherence tomography (OCT) (Optovue Inc., USA) imaging was performed in normal (230 eyes), suspect (70 eyes) and KC (120 eyes). Anterior and Bowman's surfaces were detected from the 8 radial OCT scans of 6 mm diameter. For the same surface, curvature and wavefront aberrations [6th order Zernike terms evaluated as root mean square (RMS)] were computed using a refractive index of 1.337. Discriminant performance of anterior and bowman's surface parameters were analyzed with receiver operating-characteristics (ROC). Area under the curve (AUC), Se and SP were derived.

Results:

Maximum curvature (Kmax) of anterior and bowman's surface of normal (A: 44.16±0.12D, B: 44.01±0.12D) and suspect eyes (A: 45.33±0.24D, B: 46.12±0.27D) were significantly different (p<0.01). Total RMS of A and B had an AUC of 0.92(Se: 84.9%, Sp: 80.5%) and 0.99(Se: 94.2%, Sp: 100%) between normal and suspect, 0.99(94.2%,100%) and 1(99.2%,100%) between normal and KC, respectively. On combining total RMS of A and B surfaces, AUC of 0.97(93.2%,89%) and 1.0(99.2%, 100%) was achieved, respectively. For normal vs. suspect eyes, AUC of A surface was significantly lower than AUC of both A and B combined (p=0.004).

Conclusions:

Bowman's surface parameters had better AUC than anterior surface. Combining features of anterior and bowman's surface served as a better discriminator between normal and suspect eyes. This could lead to better diagnoses of suspect eyes.

Financial Disclosure:

-