Effect of Chromatic Aberration on White-Light Contrast Sensitivity With Higher-Order Monochromatic Aberrations Correction in the Human Eye in Screen-Based Vision

Abstract

Purpose

The purpose of this study was to investigate, through theoretical modeling and experiment, the effect of chromatic aberration (CA) on white-light contrast sensitivity (CS) when higher-order aberrations (HOAs) are corrected in the human eye in screen-based vision.

Methods

Monochromatic aberrations and CA were measured in 10 subjects with normal color vision using an adaptive optics vision simulator (AOVS) under 6 mm pupil diameter. The modulation transfer function (MTF) was calculated for 6 conditions: with and without HOAs under monochromatic red, monochromatic green, and white-light, using 519 nm green light as the best-focus reference. Subsequently, white-light CS was measured under three best-focus conditions and green-light CS was measured under the green-light best-focus condition, both before and after HOAs correction.

Results

The presence of HOAs significantly improved the MTF under both green and white illumination in the presence of CA. Furthermore, the MTFs were approximately constant across red, green, and white-light conditions when HOAs were included. Under three best-focus conditions, white-light logCS demonstrated a significant decline following HOAs correction (P < 0.01). Under green-light best-focus condition, green-light logCS demonstrated a significant improvement following HOAs correction (P = 0.001).

Conclusions

The results indicate that the presence of CA in the human eye can influence the outcomes of HOAs correction in screen-based vision. Prior to implementing only HOAs correction in clinical practice, it is necessary to evaluate the effect of CA varying with individuals using an AOVS.

Translational Relevance

This research bridges the gap between optical theory and clinical practice, thereby guiding the refinement of wavefront-guided surgeries and personalized intraocular lenses.

Introduction

Under white-light illumination, retinal image quality is inevitably degraded by chromatic aberration (CA), in addition to monochromatic aberrations.^1–3 CA fundamentally arise from the dispersion properties of ocular refractive media, categorized as longitudinal chromatic aberration (LCA) and transverse chromatic aberration (TCA). The impact of CA on visual performance has been extensively investigated.^4–12 Notably, simultaneous correction of both chromatic and monochromatic aberrations has been shown to yield superior visual quality.^5–13 Whereas recent theoretical evaluations¹⁴ suggest that correcting CA could improve visual acuity by 0.2 to 0.8 lines, multiple experimental studies have found minimal functional visual improvement from CA alone.^1,7,15 The observed discrepancy between theoretical evaluations and empirical studies may arise because of the inaccurate correction of LCA and TCA in previous experiments. However, this explanation may not be comprehensive. Even if CA are precisely corrected, the interference from higher-order aberrations (HOAs; such as spherical aberration, coma, etc.) may still significantly impact the ultimate improvement in visual performance.

In 2002, McLellan et al.¹⁶ theoretically demonstrated that the interaction between CA and HOAs maintains near-constant retinal image quality across the visible spectrum. This finding provided the first theoretical confirmation of the existence of beneficial interactions between CA and HOAs. Subsequently, several theoretical studies have further elucidated the intricate interactions between chromatic and monochromatic aberrations.^17–19 Marcos et al.¹⁷ quantified the interactions between LCA and monochromatic aberrations in pseudophakic eyes using computed visual Strehl ratios. Their analysis revealed that the impact of LCA under blue light diminishes with increasing monochromatic aberrations, indicating that the interplay between monochromatic aberrations and LCA is a critical factor for optimizing visual quality in pseudophakia. Aissati et al.¹⁸ demonstrated that evaluating optical and visual quality with both natural optics and convolved images (generated using measured aberrations and Fourier optics) revealed convolution significantly degrades perceived image quality and reduces visual acuity compared to natural viewing. This discrepancy is likely caused by the convolution process eliminating the beneficial interactions naturally occurring between CA and monochromatic aberrations in the eye. Clave and Millan¹⁹ demonstrated that certain diffractive intraocular lenses preserve the critical beneficial interaction between CA and monochromatic aberrations in the human eye. Drawing from these studies, the beneficial interactions between CA and HOAs have been exclusively demonstrated through theoretical computations or indirect assessment methods. Whether the human visual system truly possesses the theorized balancing mechanism between monochromatic aberrations and CA remains to be conclusively established. Currently, no direct experimental evidence documenting their impact on visual performance has been reported in the literature.

In this study, we theoretically and experimentally investigate whether HOAs can compensate for CA-induced visual blur in screen-based vision. The interaction between HOAs and CA was assessed by analyzing the modulation transfer function (MTF) calculated from measured monochromatic aberrations and CA of the subjects. Additionally, this interaction was evaluated by comparing white-light CS before and after correction of HOAs using an adaptive optics visual simulator (AOVS).

Methods

Apparatus

This study was conducted using our laboratory-built AOVS. The schematic diagram of the reflective AOVS is presented in Figure 1. An 840 ± 25 nm superluminescent diode (SLD-371-HP1, Superlum) serves as the beacon light source. The AOVS incorporates a Badal optometer²⁰ (adjustable in 0.1 diopter [D] steps over a −4.0 D to +1.0 D range) and an astigmatism corrector²¹ (comprising two 2.0 D cylindrical lenses providing 0–4.0 D astigmatic correction) for refractive error control, along with a 69-actuator piezoelectric deformable mirror (DM; made by National Laboratory on Adaptive Optics, Institute of Optics and Electronics, Chinese Academy of Sciences) featuring a ± 2 µm stroke with 20 mm clear aperture size to correct HOAs. Ocular aberrations are measured in real-time using a Shack - Hartmann wavefront sensor (SHWS) equipped with a microlens array (ML-S400-F10; Highlight Optics Co., Ltd., Shenzhen, People's Republic of China) and a high-speed near-infrared (NIR) camera (2048 × 2048 pixels, 90 fps). The first seven orders of Zernike aberrations could be corrected effectively and higher-order terms were not amplified during closed-loop operation. The DM, SHWS, and astigmatism corrector are optically conjugated to the pupil plane. A circular aperture, likewise positioned at a conjugate of the pupil plane, is used to control the pupil size during subjective visual testing.

Figure 1.

Schematic diagram of adaptive optics visual simulator.

A real-time closed-loop control system ensures measurement reliability by automatically pausing correction during blinks or under insufficient light conditions. It is noteworthy that the spherical reflective configuration introduces additional aberrations, primarily astigmatism. During experiments, the astigmatism corrector can be used independently to correct either the inherent astigmatism of the optical path or the combined astigmatism arising from both the optical path and the subject’s eye.

Within the system, the visual stimulus delivery module comprises an OLED display (Yunnan Olightek Opto-Electronic Technology Co., Ltd., SXGA060 SC) and an achromatic lens (AL; Edmund Optics, #49-364). The measured spectra of the red (R), green (G), and blue (B) primaries, acquired using a spectroradiometer (Konica Minolta CL500A), are presented in Figure 2. The solid lines depict the raw measured spectra, whereas the dashed lines represent the same spectra weighted by the CIE 1931 standard photopic luminous efficiency function, V(λ) (source: www.cvrl.org). The peak wavelengths of the V(λ)-weighted R, G, and B spectra are 605 nm, 519 nm, and 501 nm, respectively. When the OLED display presented solid-color images of the primary colors (red, green, and blue), respectively.

Figure 2.

Normalized spectra of OLED primary colors (solid lines) and V(λ)-corrected normalized spectra (dashed lines).

The CS was measured using vertical sinusoidal gratings with Gaussian edge blur as visual stimuli in a 1.5 degrees field of view. The grating images were generated in RGB format. To achieve precise contrast adjustment, the software converted the RGB color mode to the CIE Lab color mode, keeping the a and b components constant while adjusting the L component to modify the grating contrast. The CS test used a two-alternative forced-choice paradigm. The presentation sequence in each trial was as follows¹¹: a 267-ms fixation cross signaled by a brief tone at the beginning, a 117-ms interval, a 500-ms interstimulus interval (ISI) blank, a 267-ms fixation signaled by a brief tone at the beginning, a 117-ms second interval, and blank until the response. A vertical sine-wave grating was randomly presented in one of the two intervals. The subject responded with a key press to indicate if the grating was at the first or second interval. The next trial started immediately after the response. During CS measurements, different sound prompts followed each response based on the judgment of correct or wrong. Contrast thresholds were measured with a three-down one-up staircase procedure in which three consecutive correct responses resulted in a reduction of signal contrast (C_n+1 = 0.90C_n), and one wrong response resulted in an increase in contrast (C_n+1 = 1.10C_n), converging to a performance level of 79.3% correct. In the CS test procedure, 90 trials were used to measure the contrast threshold at each spatial frequency. In addition, the starting contrast for each staircase was set slightly higher than the expected threshold based on results from pilot testing. The reciprocal of the contrast threshold was taken as the CS.

Participants

Ten undergraduates (4 men and 6 women; mean age = 20.90 ± 0.32 years; spherical equivalent = −3.48 ± 1.09 D; astigmatism = −0.45 ± 0.31 D) with normal color vision and absence of ocular pathologies were enrolled after comprehensive screening comprising: visual acuity assessment (Standard Logarithmic Chart), color vision evaluation (Ishihara pseudoisochromatic plates), autorefractometry, slit-lamp biomicroscopy, and direct fundoscopy. Exclusion criteria encompassed any detectable ocular abnormalities. Prior to participation, written informed consent was obtained with adherence to the Declaration of Helsinki under ethical approval from the Ethics Committee of Ineye Hospital of Chengdu University of Traditional Chinese Medicine (IRB Approval No. 2025YH011) and registered with the Chinese Clinical Trial Registry (ChiCTR2500109465). Prior to experimental procedures, cycloplegic mydriasis was induced in the right eye of all participants via topical administration of 0.5% tropicamide ophthalmic solution.

Measurement of Human Monochromatic and Chromatic Aberrations

In this study, monochromatic and CA data from 10 subjects under a 6 mm pupil diameter were collected using the AOVS platform. Monochromatic aberrations were measured with the SHWS. For each subject, 100 valid frames of data were collected (invalid frames due to blinks or other interference factors were excluded), and the average was taken as the final result. CA were obtained through subjective assessment methods.¹⁵ LCA was measured using the Badal optometer system. Subjects adjusted the system to achieve the sharpest view of red and green monochromatic Voronoi grid targets (diameter 1.5 degrees), respectively, and the dioptric change between the two adjustments was recorded as the LCA value. For TCA measurement, a red-green dual-color disk target (central small circle diameter 0.5 degrees; Fig. 3) was displayed on an OLED monitor. Subjects adjusted the position of the green disk (initial position randomized) using a keypad, moving it upward, downward, leftward, or rightward until they perceived the large and small cross lines (each with a line width of 3.14 arcminutes) to be fully aligned. The displacement was recorded and converted into an angular value to obtain the TCA. Each participant underwent five repeated measurements for LCA and TCA, and the average value was taken as the final result.

Figure 3.

Schematic of the visual stimulus for subjective measurement of human eye TCA.

Computation of MTF From Measured Aberrations

Based on the aberration data from each subject, this study numerically computed the horizontal MTF under six different aberration conditions at the green light (519 nm) best-focus wavelength, to align with the vertical grating assessment in the CS test. These included: (1) monochromatic red light (605 nm) without HOAs, representing the MTF of red light under diffraction-limited condition; (2) monochromatic green light without HOAs, representing the MTF considering only the CA of green light relative to red light; (3) monochromatic red light with HOAs, representing the MTF of red light in the presence of HOAs; (4) monochromatic green light with HOAs, representing the MTF with both green CA and HOAs; (5) white-light without HOAs, representing the MTF with only LCA and TCA; and (6) white-light with HOAs, representing the MTF under the combined influence of LCA, TCA, and HOAs. Given the spectral proximity between the blue and green primaries of the OLED, red-green dichromatic light was used to approximate white-light in computations. The effect of TCA was addressed using a method based on the linear shift of the point spread function. It is noteworthy that the authors measured the illuminance of the red, green, and blue primaries of the OLED display using an illuminance meter placed flush against it, which were 28.7 lx, 58.4 lx, and 25.5 lx, respectively. The green primary exhibited more than twice the luminance of the red and blue primaries. In the dichromatic MTF calculation described above, weighting factors were introduced to incorporate the influence of both the spectral sensitivity function of the human eye and the luminance imbalance between the red and green primaries of the OLED display. Finally, the MTF results from all 10 subjects under these 6 conditions were averaged for further analysis.

Experiment Procedures

An experimental investigation was conducted using the AOVS to determine whether the human visual system can perceptually benefit from the compensatory effect of HOAs against CA-induced blur. The experiment was conducted on the same 10 participants from whom aberration data had been previously acquired. The potential defensive role of higher-order monochromatic aberrations against chromatic blur was investigated by analyzing changes in white-light CS pre- versus post-HOAs correction. The experimental implementation comprised four sequential steps: (1) display of red Voronoi grid optotypes (1.5 degrees diameter; Fig. 4a) on the OLED; (2) compensation of ocular aberrations using the AOVS system, where defocus, astigmatism, and HOAs were corrected by the Badal optometer, astigmatism compensator, and DM, respectively; (3) subjective refinement of the Badal optometer setting by participants who cycled the red Voronoi grid through blur-maximum clarity-blur transitions to establish the maximum clarity position; (4) measurement of white-light CS at this optimized state across five spatial frequencies: 2, 4, 8, 16, and 24 cycles per degree (cpd).

Figure 4.

(a) The red color Voronoi grid optotypes, (b) the green color Voronoi grid optotypes, (c) the white color Voronoi grid optotypes.

In the white-light CS testing described above, the optical system was focused using red light, resulting in optical defocus of the green component. Conversely, when focusing was performed using green light, a corresponding defocus was observed in the red component. Due to the substantial luminance disparity between the red and green stimuli, these two focusing conditions are likely to yield distinct stimuli. Whether such differences significantly influence CS measurements—and consequently, the conclusions drawn from such studies—remains a pertinent and open question worthy of investigation. Therefore, we incorporated CS testing under two additional best-focus conditions to further investigate this issue. Following initial red-monochromatic best-focus testing, trials with identical configurations were subsequently conducted under green-monochromatic and white-light best-focus conditions. All three experimental arms adhered to the same testing protocol, with optotypes serving as the sole independent variable. It is important to note that across the red-monochromatic, green-monochromatic, and white-light best-focus conditions, both CA and HOAs remained consistent for each subject. Green and white Voronoi grid optotypes used in experiments are shown in Figures 4b and 4c.

In addition to optical factors, such as HOAs and CA, subjective visual performance in the human eye is also influenced by the visual neural system. Specifically, neural adaptation to HOAs may introduce confounding effects in this study. To address this issue, monochromatic green-light CS was measured under green-light best-focus conditions, both before and after HOAs correction. Measurements were conducted at two spatial frequencies: 16 and 24 cpd.

To minimize the effects of fatigue, the entire CS measurement procedure for each participant was divided into four sessions. Each session lasted approximately 1 hour and included system alignment, refractive pre-correction, and CS testing both before and after HOAs correction, as well as intermittent breaks during the measurements. The testing order for the pre- and post-HOAs correction CS assessments was counterbalanced across participants for three best-focus conditions to control for potential order effects.

Statistical Analysis

A hybrid statistical approach was used to address potential violations of parametric assumptions inherent in the repeated-measures design. Data normality was evaluated using the Shapiro-Wilk test (α = 0.05 significance threshold). A three-way repeated-measures analysis of variance (ANOVA) was conducted to assess the effects of best-focus condition, correction condition (pre-correction and post-correction) and spatial frequency on CS. Greenhouse-Geisser corrections were applied when sphericity was violated (Mauchly’s test). Significant interactions were decomposed using simple effects analyses with Bonferroni correction. For the normalization analysis of CS, data were calculated as log₁₀(X), where X denotes the measurements of CS. A P value of < 0.05 was considered statistically significant. Statistical significance was set at P < 0.05. Analytical workflows were implemented in IBM SPSS Statistics 25.0 (IBM Corp.), with visualizations generated using GraphPad Prism 9.

Results

Ocular HOAs and CA in Ten Subjects

The measurements of HOAs and CA for the right eyes of the 10 subjects are presented in the Table. The root mean square (RMS) of HOAs ranged from 0.26 to 0.59 µm, with a mean value of 0.38 ± 0.09 µm. Following the initiation of HOAs correction, the RMS of the subject’s ocular aberrations promptly decreased to below 0.1 µm, demonstrating the system’s high efficacy in correcting monochromatic aberrations. As shown in Figure 5, the time course of the RMS of typical ocular aberrations before and after adaptive optics correction is presented for subject 10, with data recorded during blinks excluded from the analysis.

Table.

Aberration Data of the Examined Right Eyes in Subjects

Subject	HOAs/um	LCA/D	TCAX/arcmin	TCAY/arcmin
S1	0.39	0.45	−0.22	0.29
S2	0.35	0.55	−0.12	0.19
S3	0.32	0.45	0.24	−0.3
S4	0.41	0.48	0.02	0.2
S5	0.35	0.44	0.56	−0.16
S6	0.59	0.50	0.64	−0.52
S7	0.41	0.52	0.07	−0.11
S8	0.42	0.52	0.37	−0.1
S9	0.27	0.49	0.09	−0.27
S10	0.26	0.50	−0.62	0.56

Figure 5.

The time course of the RMS of typical ocular aberrations of subject 10 before and after adaptive optics correction.

LCA showed limited intersubject variability, with values ranging consistently from 0.44 to 0.55 D and a mean of 0.49 ± 0.03 D. In contrast, TCA exhibited substantial variation across individuals. TCA_X values spanned from −0.62 to 0.64 arcminutes, whereas TCA_Y values ranged from −0.52 to 0.56 arcminutes. In summary, whereas HOAs and LCA remained relatively consistent within a defined range among the subjects, TCA demonstrated more pronounced interindividual differences.

MTF Computed From Measured Aberrations Data of Subjects

Figure 6 presents the averaged MTF computed from the measured aberrations of 10 subjects for a 6 mm pupil diameter. In the figure, the solid blue line represents the MTF for green light under diffraction-limited condition. The dashed blue line, indicating the MTF for green light in the presence of HOAs, shows a substantial reduction due to the impact of HOAs. The solid red line corresponds to the MTF considering only the CA of red light relative to green light, and also exhibits a significant decline resulting from LCA. The dashed red line, which denotes the MTF with both red light CA and HOAs, reveals a noticeable decrease from the diffraction-limited performance; however, it demonstrates considerable improvement compared to the MTF under CA alone. The solid black line, representing the MTF with only LCA and TCA, displays a strong reduction in optical quality, closely aligning with the monochromatic red MTF. In contrast, the dashed black line, illustrating the MTF under the combined influence of LCA, TCA, and HOAs, exhibits a significant enhancement over the solid black line. Moreover, this curve closely aligns with both the red dashed and blue dashed MTF trajectories.

Figure 6.

The averaged MTF computed from the measured wave aberrations of 10 subjects for a 6 mm pupil diameter.

Furthermore, when both red CA and HOAs are present, the MTF shows significant improvement compared to the condition with CA alone, indicating that HOAs can partially compensate for the optical degradation induced by chromatic error. The approximate constancy of MTF curves under monochromatic red, monochromatic green, and white-light conditions—when HOAs are included—further demonstrates that HOAs effectively mitigate the optical deterioration typically caused by CA. This compensatory effect is particularly evident in the condition combining LCA, TCA, and HOAs, where the MTF (represented by the black dashed line) is markedly higher than that with LCA and TCA alone (black solid line). Notably, the close agreement among the three MTF curves involving HOAs suggests that the eye’s inherent aberration structure reduces optical performance variations across different wavelengths, thereby helping to maintain consistent spatial vision under varying spectral illumination conditions. Together, these results confirm that HOAs not only partially offset blur caused by CA but also improve optical performance under polychromatic light, highlighting their essential functional role in integrating monochromatic and chromatic visual information.

Figure 7 displays the MTF computed from measured aberrations for each individual subject for a 6 mm pupil diameter. It can be observed that whereas intersubject variability in MTF exists across the six optical conditions, the overall trends remain highly consistent with the averaged MTF curve presented in Figure 6. These results indicate that, despite specific differences in individual ocular optics, the interaction between HOAs and CA exhibits a common compensatory mechanism across observers.

Figure 7.

The individual MTF computed from the measured wave aberrations of 10 subjects for a 6 mm pupil diameter.

To examine the effect of pupil size on the interaction between HOAs and CA, we also computed the MTF from the measured aberration data of 10 subjects under a 4 mm pupil condition. The averaged result is presented in Figure 8. The overall trend remains similar to that under the 6 mm condition: HOAs continue to partially compensate for the optical degradation induced by CA, although the compensatory effect is reduced compared to the larger pupil size.

Figure 8.

The averaged MTF computed from the measured wave aberrations of 10 subjects for a 4 mm pupil diameter.

CS Measurement Under Monochromatic Red-Light Best-Focus Condition

Figure 9 displays the mean white-light logCS of 10 subjects under monochromatic red-light best-focus condition, both before and after HOAs correction. A discernible change in CS was evident following correction. Statistical analysis revealed a significant overall reduction in logCS after HOAs correction relative to the pre-correction condition (F (1, 9) = 28.32, P < 0.001, partial η² = 0.759). This decrease was especially prominent within the medium-to-high spatial frequency range (8–24 cpd). At 8 cpd, a statistically significant difference in logCS was detected between pre- and post-correction (1.11 ± 0.08 vs. 0.79 ± 0.26, P = 0.002), with a mean reduction of 0.32 log units. Similarly, at 16 cpd, a highly significant difference was observed (P < 0.001), corresponding to a mean decrease of 0.31 log units. At 24 cpd, the reduction remained statistically significant (P = 0.021), with a mean decline of 0.01 log units. Significant differences in logCS across spatial frequencies were consistently evident between the two conditions (F (4, 36) = 498.00, P < 0.001, partial η² = 0.982). Furthermore, a significant interaction was also identified between spatial frequency and correction condition (F (4, 36) = 7.76, P < 0.001, partial η² = 0.775), indicating that the effect of HOAs correction on CS was modulated by spatial frequency. Specifically, at lower spatial frequencies, the correction resulted in minimal or negligible reduction in CS; however, its adverse effect became progressively more substantial with increasing spatial frequency.

Figure 9.

Pre- versus post-HOAs correction: mean white-light logCS measured under monochromatic red-light best focus.

Figure 10 presents the individual white-light logCS measured under monochromatic red-light best-focus conditions. It can be observed that although intersubject variability exists in the magnitude of logCS changes, the overall trends remain highly consistent with the averaged logCS curve shown in Figure 9. These results indicate that correction of HOAs lead to a reduction in CS in the presence of CA, suggesting a compensatory interaction between HOAs and CA that was consistently observed across all subjects.

Figure 10.

Pre- versus post-HOAs correction: individual white-light logCS measured under monochromatic red-light best focus.

CS Measurement Under Monochromatic Green-Light and White-Light Best-Focus Condition

Figures 11a and 11b show the mean logCS under monochromatic green-light best-focus condition and white-light, respectively, before and after HOAs correction. The trends observed are consistent with those under the monochromatic red-light best-focus condition, demonstrating a significant reduction in logCS following HOAs correction (green-light: F (1, 9) = 54.97, P < 0.001, partial η² = 0.859; white-light: F (1, 9) = 11.42, P = 0.008, partial η² = 0.559). This decrease was especially pronounced within the medium-to-high spatial frequency range (8 and 16 cpd), in agreement with the results under the monochromatic red-light best-focus condition. At 8 cpd, a statistically significant difference in logCS was observed before and after correction (green-light: P < 0.001 and white-light: P = 0.001), with a mean reduction of 0.33 log units (green-light) and 0.24 log units (white-light). Similarly, at 16 cpd, the difference was also significant (green-light: P < 0.001 and white-light: P = 0.011), showing a mean decrease of 0.28 log units (green-light) and 0.25 log units (white-light). In addition, no significant differences in logCS were observed both pre-(F (2, 18) = 1.78, P = 0.197, partial η² = 0.165) and post-correction (F (2, 18) = 0.478, P = 0.628, partial η² = 0.050) across all three best-focus conditions.

Figure 11.

(a) Pre- versus post-HOAs correction: mean white-light logCS measured under monochromatic green-light best-focus, (b) pre- versus post-HOAs correction: mean white-light logCS measured under white-light best-focus.

Green-Light CS Measurement Under Monochromatic Green-Light Best-Focus Condition

Figure 12 shows the mean green-light logCS at spatial frequencies of 16 and 24 cpd under monochromatic green-light best-focus condition, before and after HOAs correction. The results demonstrate significant improvement in green-light logCS at 16 and 24 cpd following HOAs correction (F (1, 9) = 20.89, P = 0.001, partial η² = 0.70). This finding confirms that the visual system retains the capacity to utilize the improved optical quality resulting from HOAs correction. Consequently, the absence of CS improvement under white-light condition cannot be attributed to an inability of the neural visual system to adapt to the altered wavefront.

Figure 12.

Pre- versus post-HOAs correction: mean green-light logCS measured under monochromatic green-light best-focus.

Discussion

This study is the first to validate, through both theoretical computation and experimental verification, that HOAs in the human eye can compensate for visual blur induced by CA The results indicate that the presence of CA in the human eye can influence the outcomes of HOAs correction in screen-based vision. Prior to implementing only HOAs correction in clinical practice, it is necessary to evaluate the effect of CA varying with individuals using an AOVS. In this study, we first measured the HOAs and CA in 10 human subjects. Based on these data, we computed the MTF under three illumination conditions—monochromatic red-light, monochromatic green-light, and white-light—both with and without HOAs. The results demonstrated that the presence of HOAs significantly enhanced the MTF under both monochromatic red and white illumination in the presence of CA. Subsequently, we assessed white-light CS in the same 10 participants using an AOVS under 3 best-focus conditions: red-light, green-light, and white-light. The results showed that white-light CS was significantly reduced after HOAs correction compared to that before correction, particularly at medium and high spatial frequencies (8–24 cpd). The experimental results are consistent with our theoretical MTF predictions, indicating that when CA remains uncorrected, HOAs are not entirely detrimental but can instead be beneficial by compensating for chromatic effects to balance visual performance.

From an optical theory perspective, the compensation of CA by the eye’s HOAs is essentially a process where two types of wavelength-dependent wavefront errors interact within the pupil region, ultimately enhancing image quality. This is exemplified by the interplay between positive spherical aberration and LCA. The LCA imposes a posterior focal shift for long-wavelength red light, which is counteracted by the anterior focal shift imposed on marginal rays by positive spherical aberration. The partial annulment of these opposing effects reduces the overall chromatic focal error, thereby concentrating the energy of the retinal blur and optimizing the point spread function for the corresponding wavelengths.

The theoretical findings of this study align with earlier predictions made by McLellan et al.¹⁶ Their results revealed that monochromatic aberrations, by counterbalancing CA, significantly enhance the potential image quality for S-cone photoreceptors and consequently maintain relatively stable retinal image quality across the entire visible spectrum. To better fit actual experimental conditions, our theoretical model incorporated luminance variations among different chromatic channels of the OLED display, in contrast to the equal-luminance assumption used by McLellan et al., which may not reflect most real-world scenarios. More importantly, this study provides experimental validation of the theoretical predictions, thereby overcoming limitations inherent in purely theoretical approaches. The work of Marcos et al.¹⁷ suggested that the impact of LCA in the blue spectrum depends largely on the magnitude of monochromatic aberrations in eyes implanted with monofocal aspheric intraocular lenses. Our results provide a plausible explanation for their conclusion.

In contrast, Yoon et al.²² reported a slight improvement in CS following HOAs correction under both monochromatic and white-light conditions, which contradicts the results presented here. This discrepancy may be explained by the specific scope of our conclusions: all participants in this study were young individuals (age = 20.90 ± 0.32 years) with relatively low levels of HOAs (RMS = 0.38 ± 0.09 µm). Thus, our conclusions are derived from this particular demographic, and their generalizability may be limited by factors such as age and aberration magnitude. Additionally, the very small sample size in Yoon et al.’s study—only two participants—may undermine the broader applicability of their findings, which could also contribute to the observed discrepancy. In 2009, Sabesan et al.²³ investigated the visual quality of keratoconic eyes after correcting HOAs. Although the study did not directly compare visual performance before and after HOAs correction, the post-correction visual acuity data indicated an improvement in white-light visual acuity. It is important to note that this study focused on patients with keratoconus, whose HOAs levels are significantly higher than those of healthy individuals. Consequently, in keratoconic eyes, HOAs and CA may not effectively compensate for each other as they do in healthy eyes. We hypothesize that beyond a certain threshold of HOAs (as seen in conditions like keratoconus or cataracts where HOAs are substantial), the compensatory visual mechanism may operate differently. This aspect will be explored in future research.

In this study, we examined the impact of illuminance imbalance among the three primary colors in OLED displays by evaluating CS under three best-focus conditions: red monochromatic, green monochromatic, and white-light. The results revealed no significant differences in CS before and after HOAs correction across all conditions. These findings suggest that, within the current experimental framework, luminance-induced variations in visual stimuli did not substantially influence CS measurements, and thus did not markedly alter the compensatory relationship between HOAs and CA. The artificial lighting and display technologies—such as filament lamp, LED lamp, computer monitor, OLED display, smartphone display, and so on—are being increasingly adopted. It is therefore essential to further investigate how these light sources affect the interaction between CA and HOAs in the human eye. Such insights are critical for advancing aberration correction strategies. However, their spectral characteristics remain fundamentally distinct from those of natural light. Under natural illumination, the broader spectral width and more pronounced CA may lead to more complex interactions among aberrations. This study serves as an initial step in this research direction, and subsequent work should use more comprehensive and rigorous methodologies to clarify the underlying mechanisms.

Moreover, in the present study, subjective visual performance measurements were influenced not only by optical factors including monochromatic and CA, but also by neural compensatory mechanisms. Previous research^24–28 has established that the human visual system exhibits partial neural compensation for both monochromatic aberrations and CA. To investigate whether the decline in white-light CS is mediated by neural adaptation mechanisms, we performed green-light CS measurements under monochromatic green-light best-focus condition. Contrary to the predicted decrease in CS following HOAs correction – which would be expected if neural adaptation were the dominant mechanism – we observed a significant improvement in green-light CS after correcting HOAs under monochromatic green-light best-focus condition. This demonstrates that the interaction between HOAs and CA, rather than neural adaptation mechanisms, constitutes the dominant factor underlying the observed differences in CS changes.

As the clinical application of HOAs correction technologies continues to expand, their efficacy in meaningfully improving visual outcomes remains a topic of ongoing debate. Evidence indicates that wavefront-guided refractive surgery does not consistently surpass conventional approaches in performance,^29–32 and no clear consensus exists regarding whether spherical aberration should be fully corrected after aspheric intraocular lens (IOL) implantation.^33–39 These observations may be partially attributed to the inherent limitations of current techniques in achieving complete HOAs correction. However, the present study reveals that the presence of CA in the human eye can influence the outcomes of HOAs correction—suggesting it as a potential contributing factor to the aforementioned clinical controversies. Therefore, prior to implementing only HOAs correction in clinical practice for vision tasks dominated by screen use, it is necessary to evaluate the effect of CA varying with individuals using an AOVS.

This study utilized cycloplegia to immobilize accommodation and set a 6 mm pupil diameter. These conditions facilitated precise measurement of the interaction between HOAs and CA in a controlled environment, yet they also differ from natural viewing states where accommodative fluctuations and dynamic pupil changes are present. In younger individuals, accommodative fluctuations may modulate the interaction between HOAs and CA. Future research could further explore this interaction under conditions that allow for natural accommodation and pupil dynamics.

This study has certain limitations. It should be noted that all subjects were university students and relatively young, which corresponds to generally lower magnitudes of HOAs. Because HOAs tend to increase with age^40,41 and certain ocular pathologies,^42–44 whether the conclusions of this study apply to populations with abnormally high levels of HOAs requires further analysis and validation.