Just-noticeable differences for wavefront aberrations obtained with a staircase procedure

To explore the magnitude of aberrations that alter the subjective visual quality measurably. The suitability of a staircase procedure for threshold determination was investigated.

Spectacle correction (correction A) was achieved in 28 healthy eyes by an adaptive optics set-up, and additional astigmatism, coma, or trefoil was induced (correction B). Subjects observed a radial test pattern with noncyclopleged eyes using an artificial 5-mm pupil. They compared corrections A and B and identified the subjectively better correction. The magnitude of the aberration for each presentation varied according to the "Best PEST" staircase procedure. Just-noticeable difference (JND) was found after 30 presentations. Participants were divided into three groups, and the experiment was repeated under the following conditions: without modifications (n = 9), with a letter chart (n = 10), and with full correction of all aberrations (n = 9).

Median JND was 0.091 μm for astigmatism, 0.059 μm for coma, and 0.108 μm for trefoil, with a large interindividual spread. Repeatability coefficient across the three aberrations was 0.095 μm. A significant correlation was found between JNDs and the residual aberrations (without the generated aberration) in astigmatism and coma. Differences between radial test pattern and letter chart and between spectacle correction and full correction were not significant.

JNDs were lower for coma than for astigmatism and trefoil, depending on the magnitude of other residual aberrations, and showed large variations among subjects. Thus, the decision to correct more pronounced aberrations (eg, by wavefront-guided LASIK) should not depend on wavefront measurement results alone. A staircase procedure for determination of JNDs provides repeatable results.