[Discussion]: The study was focused on the above mentioned 8 eyes, which sho-wed differences greater than 1 D between subjective and objective equivalent sphere, Se. Se was obtained subjectively by standard subjective refraction and also by a more precise custom Badal system. For the objective refraction we used retinoscopy, autorefracto-meter (Canon T1000) and aberrometer (irx3, Imagine Eyes). In what follows we present one example of bifocal eye, which showed the maximum discrepancy of nearly 2.5 D between subjective and ob-jective refraction. This case is especially interesting, since there are two different possible explanations for such a large difference. The left panel of Fig. 1 shows the distribution of equivalent sphere across a 4.9 mm pupil diameter, showing values ranging from -5 to -9 D. The right panel shows the frequency histogram of that distribution under three hypothetical situations. Red line corresponds to the histogram of the plain aberrometric measure which implicitly assumes a perfect circular pupil with uniform transmission. This histogram shows two peaks separated by more than 2.5 D, which is indicative of a bifocal wavefront. Nevertheless, the peak on the right is domi-nant as it shows Se = -5.8 D. The blue histogram simulates the effect of pupil vignetting by the upper eye lid. This simulation was inclu-ded, since this phenomenon was actually observed in this subject. For that purpose we assumed that the area of the pupil above the das-hed line was blocked by the eye lid. The resulting (blue) histogram is totally different, now showing a single peak, which is much closer to the value of the subjective refraction (Se=-8.2 D). The green histogram represents a second possible explanation. It was computed by weighting the data points according to the Stiles-Crawford effect (SCE). Here we used standard nominal values, ex-cept for the fact that we shifted the SCE peak downwards (its positi-on is indicated by the cross on the left panel.). As expected, the decentered SCE yields an intermediate situation: The two peaks remain but their relative heights changed so that now the dominant peak is the left one (Se =-8.2 D). The other cases analyzed so far exhibit two or even three peaks, which means that there are two or three possible refractions. One of these peaks is always close to the subjective refraction. Their relative heights depend on the position of the SCE peak.

Several examples of eyes with multifocal wavefronts are analyzed by comparing subjective and aberrometric refraction. These multifo-cal eyes are analyzed using a generalized refractive error sensing method. Our results suggest an important effect on the refraction caused by inhomogeneous pupil transmittance (SCE or vignetting.). [Introduction]: In some eyes subjective refraction may differ more than 1 D from objective refraction obtained by aberrometry or autorrefracto-metry. Usually these eyes present wavefront errors larger than ave-rage (in particular coma and/or spherical aberration). Here, our goal was to assess the hypothesis that these important discrepancies in these highly aberrated eyes might be explained by their multifocal properties. In normal eyes, the prevalence of these cases is not high. We identified 8 out of 178 eyes (~5%) from a previous study. However the prevalence in post-surgical eyes is expected to be much higher. The analysis was based on refractive error sensing, which was generalized to account for multifocal wavefronts, as well as to include the effect of possible irregular and/or non-uniform transmission pupils (Stiles-Crawford effect, etc.).

[Conclusions]: A small percentage of normal eyes show bifocal wavefronts, with power differences greater than 1 D (or even 2.5 D in some cases), which are typically associated to high values of HOA. These bifocal eyes often show important discrepancies between subjective and objective refraction. Our results suggest that these discrepancies may appear because aberrometric methods ignore the inhomogene-ous effective pupil transmittance. The generalization of refractive error sensing proposed here seems suitable to obtain the refraction in these multifocal eyes. It may be especially relevant for postsurgi-cal eyes, which often present high levels of HOA, and hence may show multifocal wavefronts.

Trabajo presentado al 6th EOS Topical Meeting on Visual and Physiological Optics celebrado en Dublin (Irlanda) del 20 al 22 de agosto de 2012.

Supported by grant FIS2011-22496 (Spain).

Peer reviewed