Abstract
Purpose:
Algorithm to read simulated visual acuity from both monofocal and multifocal eye models with varying levels of defocus.
Methods:
We developed an algorithm capable of correlating individual letters of an aberrated letter chart with an unaberrated chart. We correlated the letter charts generated from both monofocal and multifocal eye models with varying amounts of defocus. Two metrics were developed, one for each type of model. For the monofocal model, a contrast-based metric was used for detection of a readable letter. For the multifocal model, a weighted metric of both contrast and enhanced edge detection was used. These metrics allow for the detection of the presence of a readable letter at a given location in the aberrated letter chart with respect to the unaberrated chart. For each metric, a threshold was found representing the cutoff in readability of the chart. The monofocal contrast-based metric was verified using clinical visual acuity measurements for defocus ranging from −2.0D to +2.0D. In addition, simulated blurred eye charts were created and readers with normal acuity read these charts to determine a standard observer visual acuity. Thresholds were chosen to give the best match between clinical/standard observer visual acuity data and simulation for a 3.5mm pupil size.
Results:
For the monofocal case with a 3.5mm pupil, our metric predicts the standard observer visual acuity for defocus values over the range ±2.0D well, but consistently over predicts the clinical visual acuity within the range -0.75D to +2.0D. For the multifocal case with a 3.5mm pupil, the algorithm adequately predicts the standard observer visual acuity except for two regions: for myopia worse than −4.0D; and defocus values near −2.0D between the distance and near portions of the multifocal.
Conclusions:
We demonstrated an algorithm for estimating the visual acuity of letter charts generated with varying defocus for both monofocal and mulitfocal eye models. With a simple contrast based metric, monofocal visual acuity can be estimated for defocus values outside the range ±0.5D. Inside the range ±0.5D, the model over-predicts visual acuity likely due to the model not accounting for limitations of aberrations, scatter and sampling in the eye. With a weighted contrast and enhanced edge detection metric, multifocal standard observer visual acuity can be well estimated over the range −4.0D to +1.5D.
Keywords: 754 visual acuity •
626 aberrations •
473 computational modeling