Optom Vis Sci . 2026 Mar;103(3):e70021. doi: 10.1002/ovs2.70021. ABSTRACT PURPOSE: (1) To investigate if binocular cross-correlation metrics derived from the aberrated monocular retinal images can explain the deterioration of stereoacuity following keratorefractive surgeries. (2…
Optom Vis Sci. 2026 Mar;103(3):e70021. doi: 10.1002/ovs2.70021.
ABSTRACT
PURPOSE: (1) To investigate if binocular cross-correlation metrics derived from the aberrated monocular retinal images can explain the deterioration of stereoacuity following keratorefractive surgeries. (2) To investigate if the quality of binocular metrics improves with improved monocular retinal image quality and pupil miosis.
METHODS: High-order aberrations (HOAs) were obtained over 6 mm pupil diameter from 99 young adults undergoing photorefractive keratectomy (PRK) (n = 26), laser-assisted in situ keratomileusis (LASIK) (n = 38), and ReLEx small-incision lenticule extraction (SMILE; n = 35) for myopia correction. Point spread functions derived from HOAs were convolved with random-dot stereo pairs and cross-correlated to obtain four binocular metrics-signal width, signal height, signal-to-noise ratio, and height-to-width ratio-that signified the quality of the disparity signal. These metrics were derived pre-operatively and at 1-week, 1-month, 3-month, and 6-month after surgeries, and correlated with the subject's stereoacuity for 6 mm pupil diameters. Binocular metrics were also derived for 4 mm pupil diameters and compared with the 6 mm data.
RESULTS: Larger interocular differences in retinal image quality led to reduced disparity signal quality and less optimal values of the binocular metrics. These trends were strongly correlated with the subject's stereoacuity across refractive surgeries (ρ ≥ 0.71; p < 0.001). Relative to pre-operative values, post-operative metrics worsened at 1-week followed by a recovery period (p < 0.001). All post-operative metrics for PRK and LASIK surgeries were similar (p ≥ 0.67), but worse than SMILE surgery (p < 0.001). The binocular metrics derived for the 4 mm pupil diameters were superior to those of the 6 mm pupil diameters (p < 0.001).
CONCLUSIONS: Binocular cross-correlation metrics derived from the aberrated monocular retinal images signify the strength of the disparity signal, and capture much of the stereoacuity losses following keratorefractive surgeries for myopia. Binocularity may be better preserved after SMILE surgery owing to the smaller interocular mismatches in optical quality, relative to PRK and LASIK surgeries for comparable pupil diameters.
PMID:42020351 | DOI:10.1002/ovs2.70021