Predicting visual performance from optical quality metrics in keratoconus
Related Experiment Video
Updated: Jun 17, 2026
Full-Field Optical Coherence Microscopy for Histology-Like Analysis of Stromal Features in Corneal Grafts
Published on: October 21, 2022
Paul Schoneveld1, Konrad Pesudovs, Douglas J Coster
1NH&MRC Centre for Clinical Eye Research, Department of Ophthalmology, Department of Ophthalmology, Flinders Medical Centre and Flinders University of South Australia, Bedford Park, South Australia, Australia.
Purpose:
The aim was to identify optical quality metrics predictive of visual performance in eyes with keratoconus and penetrating keratoplasty (PK) for keratoconus.
Methods:
Fifty-four participants were recruited for this prospective, cross-sectional study. Data were collected from one eye of each participant: 26 keratoconus, 10 PK and 18 normal eyes: average age (mean +/- standard deviation) 45.2 +/- 10.6 years and 56 per cent female. Visual performance was tested by 10 methods including visual acuity (VA), both high and low contrast (HC- and LC-) and high and low luminance (LL-), and Pelli-Robson contrast sensitivity, all tested with and without glare. Corneal first surface wavefront aberrations were calculated from Orbscan corneal topographic data using VOLPro software v7.08 (Sarver and Associates) as a tenth-order Zernike expansion across three, 4.0 mm and 5.0 mm pupils and converted into 31 optical quality metrics. Pearson correlation coefficients and linear regression were used to relate wavefront aberration metrics to visual performance.
Results:
Visual performance was highly predictable from optical quality with the average correlation of the order of 0.5. Pupil fraction metrics (for example, PFWc) were responsible for all of the highest correlations at large pupils for example, with HCVA (r = 0.80), LCVA (r = 0.80) and LLLCVA (r = 0.75). Image plane metrics, derived from the optical transfer function (OTF) were responsible for most of the highest correlations at smaller pupils for example, volume under the OTF (VOTF) with HCVA (r = 0.76) and LCVA (r = 0.73).
Conclusions:
As in normal eyes, visual performance in keratoconus was predicable from optical quality; albeit by different metrics. Optical quality metrics predictive of visual performance in normal eyes, for example, visual Strehl, lack the dynamic range to represent visual performance in highly aberrated eyes with keratoconus. Optical quality outcomes for keratoconus could be reported using many different metrics, but pupil fraction metrics, for example PFWc, perform best for highly aberrated eyes.
