Population-based simulation using image quality metrics to predict visual acuity in pseudophakic patients implanted with trifocal IOLs

Abstract

Purpose : Visual acuity is an essential clinical endpoint to assess the effectiveness of intraocular lenses (IOLs) in clinical trials. Recent innovative trifocal IOLs have been shown to provide patients with great visual acuity at distance, intermediate and near. Population-based simulation of visual acuity at a range of defocuses (i.e., defocus curve) can help to optimize IOL optical design and guide clinical study design.

Methods : One-hundred virtual eyes have been generated using two-surface reduced eye model. Monte-Carlo approach was adopted by iterating different corneal power and aberration, anterior chamber depth and pupil size. The IOL surface was extracted from normative lens design in grid-sag format. Two major metrics, MTF area from 0 to 50 c/mm (MTFa50) and light-in-the-bucket (LIB), were calculated at varied defocuses from -3.5 D to +1.0D. The metric-visual acuity correlation was established using existing clinical data of a bifocal IOL. The correlation functions (MTFa50: a*x^b+c; LIB: a*log(LIB)+b) were applied to the new trifocal IOL to simulate its visual acuity and defocus curve.

Results : The LIB has shown better correlation with clinical binocular visual acuity than MTFa50 (R²=0.92 vs 0.64). The correlation functions were 6.7*x^(-0.047)-5.2 for the MTFa50 and -0.24*log(LIB)-0.20 for the LIB. Similar to the clinical binocular visual acuity, both MTFa50 and LIB metrics predicted the trifocal lens could provide visual acuity better than 0.12 logMAR from 0D (distance) to -2.5D (near) (Figure 1). Simulated difference (MTFa50-LIB) ranged from -0.07 to 0.03 logMAR, all within 1 line. MTFa50 predicted defocus curve was more smooth and monotonic without a distinct secondary peak. Bland-Altman analysis suggested that MTFa50 tended to predict worse visual acuity than LIB when the visual acuity was good and to predict better visual acuity when the visual acuity was poor (Figure 2).

Conclusions : The predicted visual acuity using population-based MTFa50 and LIB metrics have been generally consistent with each other. Calculating MTF area metric only up to 50 c/mm rather than traditional 100c/mm may act as a low-pass filter in predicting visual acuity and defocus curve.

This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.