Accommodation to Monochromatic and White-Light Targets

Transcription

1 Accommodation to Monochromatic and White-Light Targets Karan R. Aggarwala, Sujata Nowbotsing, and Philip B. Kruger Purpose. The objective of the current study was to compare accommodation to targets illuminated with monochromatic light from different regions of the visible spectrum with accommodation to white-light targets. Methods. One of 10 narrow-band interference filters (430, 450, 470, 500, 530, 550, 570, 590, 630, and 670 nm) was used to produce monochromatic light from a tungsten-halogen source to illuminate a Maltese cross-target in Maxwellian view. Luminance of each monochromatic light was matched by minimum border photometry against a standard white light (3000 K) that was maintained at 200 cd/m 2. Chromatic difference of focus of the eye was minimized for all monochromatic targets by the use of an achromatizing lens. A white-light target also was used, and the subject's eye was achromatized or the eye had normal chromatic aberration. The target was moved sinusoidally toward and away from the eye at a temporal frequency of 0.2 Hz over a 1 D amplitude (peak to peak). Accommodation was monitored continuously by an infrared recording optometer, and responses were Fourier analyzed to obtain gain arid phase lag at the temporal frequency of stimulation. Results. Accommodative gain was highest and phase lag was smallest when the target was illuminated by white light in the presence of normal chromatic aberration. The achromatized white-light gain of accommodation was statistically similar to the gain for monochromatic targets, indicating that the presence of chromatic aberration facilitates accommodation. Significant intersubject variability was present in the accommodative tracking ability to monochromatic targets. Conclusions. Accommodation to monochromatic targets is not as accurate as accommodation to a white-light target, and this effect is related to the presence of ocular longitudinal chromatic aberration for the white-light target. Invest Ophthalmol Vis Sci. 1995;36: A he primate eye has the remarkable ability to change the dioptric power of its crystalline lens to focus a target of interest (accommodation). The monocular focusing ability of the eye is guided in part by perceptual mechanisms that encode depth 1 " 4 and in part by optical (dioptric) stimuli. 5 The influence of cognitive demand 6 ' 7 and volition 8 " 10 on the accommodation response also has been demonstrated. In addition, the dioptric stimulus for accommodation is influenced by the initial contrast of the target 1112 and by the spatial frequency content of the target. 13 " 15 From the Schnurmacher Institute for Vision Research, Slate University of New York, State College of Oplometry, New York, New York. Supported in part by a Schnurmacher Foundation Grant and by awards from the National Eye. Institute (F32-EYO6403, EYO5901). Submitted for publication May 11, 1995; revised August 15, 1995; accepted August 16, Proprietary interest category: N. Refjrint requests: Karan R. Agganuala, SUNY-State College of Optometry, 100 East 24th Street, Room 1334, Nexu York, NY The eye has substantial longitudinal chromatic aberration, and short-wavelength light focuses anteriorly compared to light of longer wavelength. l617 Fincham 18 was the first to suggest with supporting evidence that longitudinal chromatic aberration (LCA) acts as a stimulus to accommodation. He showed that accommodation is impaired in most subjects when LCA is reduced by an achromatizing lens or a monochromatic target. In monochromatic sodium light (589 nm) 18 and in narrow-band 590 nm light, 19 " 21 the reflexive accommodation response to a dioptric stimulus is not as accurate as the response in broad-band white light. The current study aims to investigate whether a similar debilitation of accommodation occurs in monochromatic light from other parts of the visible spectrum. Targets with restricted spectral content may serve as adequate stimuli for accommodation when the ac- Investigative Ophthalmology & Visual Science, December 1995, Vol. 36, No. 13 Copyright Association for Research in Vision and Ophthalmology 2695

2 2696 Investigative Ophthalmology & Visual Science, December 1995, Vol. 36, No. 13 S L1 L2 L3 T L4 C L5 P1 T L6 FIGURE l. Badal optical system for presenting targets and for stimulating accommodation. Illumination optics are depicted as dashed lines, and solid lines represent target optics. Monochromatic light could be produced by introducing interference filters between lenses LI and L2. An achromatizing lens (A) eliminated chromatic difference of focus of the eye for the monochromatic conditions and for one whitelight condition. commodation response is supplemented substantially by voluntary effort 22 or by anticipation. 23 Stationary targets facilitate voluntary focusing, and for static targets, spectrally band-limited targets may appear to be adequate stimuli for accommodation However, when a step change in focus is presented to the eye, the accommodation response is completed in less than 1 second 2 ' 1 ' 27 and such responses are more likely to be reflexive rather than voluntary. Reflexive focusing occurs in response to the introduction of low-power lenses 1 and to changes in target position in a Badal optical system. 28 Moving targets were used in the current study to reduce the influence of voluntary accommodation. MATERIALS AND METHODS The optical system used for presenting targets and stimulating accommodation is depicted in Figure 1. Light rays of the illumination system are illustrated by the dashed lines, and the path of light from the target is represented by the solid lines in Figure 1. Illumination Optics Light from a tungsten-halogen source (S) was used to illuminate the target-slide (T) using a system of three lenses (LI, L2, and L3). One of 10 interference filters (430, 450, 470, 500, 530, 550, 570, 590, 630, and 670 nm) were introduced in the parallel beam of light between lenses LI and L2. Lens L4 formed an image of the light source in the plane of a circular aperture (C). An achromatizing lens (A), mounted in a small filter-wheel, could be introduced in the same plane as the aperture to neutralize the longitudinal chromatic aberration of the eye. The achromatizing lens was a doublet lens, and it undercorrected LCA by approximately 0.3 D. Transverse chromatic aberration at the fovea of the human eye was not altered in the presence of the lens, 21 and the lens did not alter contrast sensitivity of the eye. 21 The aperture (C), achromatizing lens (A), and the tungsten source (S) were imaged in the plane of the subject's pupil by lenses L5 and L6 in Maxwellian view. 29 ' 30 The image of the circular aperture served as a 3-mm artificial pupil in the subject's eye. Target Optics The target consisted of four illuminated radial wedges presented in the form of a Maltese cross 19 on a dark background. The Maltese cross was illuminated either by white light or by monochromatic light from different parts of the visible spectrum (430, 450, 470, 500, 530, 550, 570, 590, 630, and 670 nm). The white light was produced by a tungsten-halogen source (3000 K), and interference filters (10 nm bandwidth at halfpeak transmittance) were used to generate the monochromatic lights. The spectral distribution of these monochromatic lights was similar to a narrow-band gaussian, centered at the wavelength of peak transmittance. Light from the target was collimated by lens L4 and focused by lens L5 to form an aerial image of the target (T'). The intervening mirrored prisms (PI and P2) were used to alter the optical path length so that the aerial target (T') could be moved toward or away from the Badal lens (L6) to stimulate accommodation. Prism PI was fixed, whereas prism P2 was mounted on a ball-slide and driven by a computer-controlled servo motor. The achromatizing lens was aligned and positioned within the optical system using a He-Ne laser, and the lens was imaged, along with an artificial pupil, in the plane of the subject's entrance pupil. This type of imagery eliminates the wavelength-dependent magnification that can occur when lenses are placed in the spectacle plane rather than in the nodal plane. 31 The eye was carefully aligned with the axis of the optical system by aligning the first Purkinje image of the eye with the cross-hairs of a telescope. A bite-plate assembly and forehead rest were used to stabilize the subject's head, and steady fixation was maintained during experimental trials. Infrared Optometer Accommodation was monitored continuously by a high-speed infrared recording optometer at a frequency of 100 Hz. 32 The optometer operates off a hot mirror, and its output is a voltage that varies linearly with accommodation. This voltage was digitized and scaled according to a calibration procedure conducted at the start of each experimental session. The optometer output is unaffected by small eye movements (up to 3 from fixation), and it registers the

3 Accommodation to Targets 2697 S1 STIMULUS 570 nm z O a o o 450 nm / V W A " / ^ ^ ^ 470 n 500 nm 590 nm 630 nm 670 nm 530 nm NOR FIGURE 2. Accommodative tracking of one subject (SI) to sinusoidal target motion of 1 D peakto-peak amplitude (stimulus). Accommodative tracking to a white-light target with normal chromatic aberration (NOR) appears less erratic than other conditions of illumination. 550 nm TIME (sec) ACH TIME (sec) natural high-frequency oscillations of accommoda- 33,34 tion. Subjects The research was approved by the Institutional Review Board at the SUNY-College of Optometry. Informed consent was obtained, and subjects were paid for their participation. Tenets of the Declaration of Helsinki were followed. The six subjects had good corrected visual acuity (20/20 or better), and trial lenses were placed close to the eye (in a lens holder aligned with the optical system) to compensate for refractive error. Subjects were screened for gross red-green vision defects using Ishihara plates (Graham-Field, Hauppage, NY) illuminated by a Macbeth lamp (Richmond Products, Boca Raton, FL). Procedures Luminance of the monochromatic lights was matched by minimum border photometry 35 to a white light that was maintained at 200 cd/m 2 as measured by a Pritchard photometer (Photo Research, Chatsworth, CA). The 10 monochromatic targets were presented in random order through a specially designed achromatic doublet that focused all wavelengths in approximately the same plane to minimize the longitudinal chromatic aberration of the eye. For the white-light target, the eye was either achromatized by the doublet lens, or the lens was absent and the eye had normal chromatic aberration. A Badal optical system was used to present the target, and dioptric vergence could be altered without changing brightness, size, or lateral position of the target The target was moved sinusoidally toward and away from the eye at a temporal frequency of 0.2 Hz, through a range of 1 D. The nonfixating eye was covered by an eye patch, and subjects were instructed to concentrate their attention at the center of the target. Each accommodation trial lasted approximately 40 seconds, during which the target moved through a range of 1 D toward and away from the eye in eight sinusoidal cycles at 0.2 Hz. The 10 monochromatic conditions (430, 450, 470, 500, 530, 550, 570, 590, 630, and 670 nm) and two white-light conditions (normal LCA and achromatized) were presented in random order, with short rest periods between trials. Five such experimental sessions (each lasting approximately 1 hour) were conducted for every subject.

4 2698 Investigative Ophthalmology & Visual Science, December 1995, Vol. 36, No. 13 A Fast Fourier Transform was performed on the data from each trial to obtain amplitude and phase of the focusing response at the temporal frequency of the target motion (0.2 Hz). Blinks were removed automatically before Fourier analysis. 15 Gain (response amp/stimulus amp) and phase lag (time delay) of accommodation were computed for each trial, and diese served as an index of accommodative performance. A vector average was computed for the gain and phase data from the five trials of each subject. In the vector-average procedure, gain and phase were treated as components of a single vector quantity. Gain was regarded as the magnitude of the vector, whereas phase represented the direction of the vector in a polar coordinate system. Vector addition was performed, 37 leading to a resultant vector with its own magnitude and direction. Finally, the vector average was obtained by dividing the magnitude of the resultant vector by the number of vectors (gain and phase pairs) forming the sum. RESULTS Raw accommodative tracking data for one subject (SI, 28 years of age) to 10 monochromatic targets (430, 450, 470, 500, 530, 550, 570, 590, 630, and 670 nm) and two white-light targets (normal LCA and achromatized) are presented in Figure 2. The mean stimulus to accommodation was 2 D, and die target moved sinusoidally ±0.5 D widi regard to die mean level (top traces of Figure 2). When the target was illuminated by short-wavelengdi light (430, 450, and 470 nm), the accommodation of subject SI was characterized by significant oscillations and erratic accommodative behavior. Subject SI accommodated somewhat better to mid-spectral and long wavelengdi targets than to short-wavelengdi targets. The last two traces of Figure 2 show diat die accommodation responses of subject 51 to white-light targets were relatively free of highfrequency noise or low-frequency drift compared to most odier monochromatic targets. Raw data from another subject (S2, 24 years of age) are presented in Figure 3. Unlike SI, subject 52 accommodated reasonably well to short-wavelength targets but responded poorly to long-wavelength targets. The response of subject S2 to mid-spectral targets was similar to die achromatized white-light response. For subject S2, accommodative tracking to a whitelight target improved substantially when die normal longitudinal chromatic aberration of the eye was restored (NOR, Fig. 3). Figures 4, 5, and 6 show vector-averaged (five trials) gain and phase lag of the accommodation response at the temporal frequency of stimulation (0.2 Hz) for pairs of subjects. Subjects who had a similar characteristic decline of gain in a particular region of the visible spectrum were paired for comparison. Solid bars represent accommodation to the normal LGA white-light target, single-hatched bars are for an achromatized white-light target, and cross-hatched bars represent accommodation to monochromatic targets in ascending wavelength order from left to right (430, 450, 470, 500, 530, 550, 570, 590, 630, and 670 nm). From Figure 4 it is clear that for subjects SI and S5, gain to long-wavelength targets (590, 630, and 670 nm) was higher than the gain to short wavelength targets (430, 450, and 470 nm). Both subjects (SI and S5) showed substantially reduced gain and a somewhat increased phase lag to the 450-nm target. Highest gain and smallest phase lag were obtained when the target was white with the eye's normal chromatic aberration present (solid bars). In Figure 5, data are presented from two subjects who accommodated poorly to long wavelengdi targets (630 and 670 nm) but fairly well to short wavelength targets (430, 450, and 470 nm). For subject S2, a decline in phase lag is observed for mid-spectral targets (500, 530, and 550 nm), and, for subject S4, a slighdy increased gain is evident for mid-spectral stimuli. In general, subject S4 accommodated well to most monochromatic targets, and die average gain (five trials) to the 550 nm condition (tallest cross-hatched bar, Fig. 5) was similar to diat obtained for the normal whitelight condition (solid bar). For both subjects of Figure 5 (S2 and S4), an achromatized white-light target was not as good a stimulus to accommodation as a whitelight target widi normal chromatic aberration. For subjects S3 and S6 (Fig. 6), a dramatic loss of accommodative tracking ability occurred for longwavelength stimuli. A modest decline in accommodative performance of S3 and S6 also was evident at short wavelengths. These two subjects were different from the other four subjects (SI, S2, S4, S5) in the degree to which the achromatized white-light target (single hatched bars) debilitated accommodation. Analysis of variance for a repeated measures experimental design (n= 6) showed significant intersubject variability (P 0.001), and a significant treatment effect (P 0.001) also was present. A multiple comparison test (Tukey, Honestly Significant Difference) 38 was conducted to search for significant differences between the mean accommodative gains for the 12 conditions (10 monochromatic and 2 white). The 10 monochromatic conditions and the one achromatized white-light condition were found to be statistically similar to each other, but the white-light condition with normal longitudinal chromatic aberration was significandy different from the remaining 11 conditions (P 0.01). DISCUSSION The data from the current study suggest that targets illuminated by narrow-band light (10-nm bandwidth),

5 Accommodation to Targets 2699 S2 STIMULUS STIMULUS 430 nm 570 nm z o 590 nm a o 470 nm 630 nm FIGURE 3. Accommodative tracking of another subject (S2) to sinusoidal target motion. This subject showed high-amplitude tracking to a white-light target with normal chromatic aberration (NOR). Focusing responses to the achromatized white-light target (ACH) and to most of the 10 monochromatic conditions (430 to 670 nm) were impaired compared to NOR O o 500 nm 530 nm 550 nm TIME (sec) 670 nm NOR ACH TIME (sec) as well as achromatized white-light targets, do not stimulate the eye to accommodate as accurately as normal white-light targets. In other words, white-light targets provide a good stimulus to accommodation only when the normal LCA of the eye is present. The accommodative tracking response for targets of various wavelengths varies among individuals. Some subjects accommodated poorly to short-wavelength targets (SI and S5), whereas others accommodated poorly to long-wavelength targets (e.g., S2, S3, S4, S6). In midspectral illumination, all subjects except subject S2 accommodated as well as or better than in short-wavelength or long-wavelength illumination. Intersubject Variability Significant intersubject variability was present, and each subject seemed to accommodate poorly in a particular region of the visible spectrum. A closer analysis of individual data suggests that subjects fall into three categories or response types. In Figure 4, both subjects (SI and S5) accommodated best in long-wavelength and mid-spectral light, but accommodation was poor for short-wavelength targets. The mean monochromatic gain (cross-hatched bars) for the two subjects of Figure 4 declined by 34% for SI and by 38% for S5 when compared to the mean accommodative gain to a white-light target with normal longitudinal chromatic aberration (solid bar). The second category of focusing response to monochromatic and white-light targets is depicted in Figure 5. These subjects (S2 and S4) accommodated more accurately to short-wavelength and mid-spectral targets, but a modest decline in gain and increase in phase lag was present for long-wavelength targets. The mean decline in gain from monochromatic (crosshatched bars) to white with normal LCA (solid bar) was small for S4 (18%) but larger for S2 (44%). Finally, Figure 6 presents two subjects who accommodated modestly to mid-spectral targets and showed a slight decline in focusing accuracy for short-wavelength targets and a marked loss of accommodative tracking ability to targets illuminated by long-wavelength light. These two subjects (S3 and S6) generally focused poorly in monochromatic light compared to white light with normal LCA. The average monochromatic gain of S3 was substantially smaller (by 79%) than this subject's gain to a white-light target with LCA. For subject S6, the decrease in gain from mono-

6 2700 Investigative Ophthalmology & Visual Science, December 1995, Vol. 36, No. 13 S1 S5 D CO Q_ 150 -_ 120 -J 90 " A 60 - A rl jj JL 30 n n -C_2_ > > C > x x ; i B FIGURE 4. Gain and phase lag of accommodation of two subjects (SI and S5) who accommodated poorly to short-wavelength targets. From left to right, the cross-hatched bars represent: 430, 450, 470, 500, 530, 550, 570, 590, 630, and 670 nm illumination conditions. The solid bar represents a white-light target with normal chromatic aberration (NOR), and the single-hatched bar (extreme right) depicts a white-light target without chromatic aberration (ACH). Error bars portray 1 SSEM response for five trials. chromatic (cross-hatched bars) to normal white (solid bar) was somewhat less dramatic (57%) than for subject S3. Longitudinal Chromatic Aberration In the current experiment, a white-light target viewed through an achromatizing lens debilitated accommodation compared to a white-light target with normal LCA. The reduction in gain from normal LCA white to achromatized white-light targets (NOR-ACH) was similar to the reduction in gain from normal LCA white to the average monochromatic gain (NOR- MONO) for the group of six subjects (Tukey HSD multiple comparison test). These results suggest that the LCA of the eye provides an important dioptric stimulus for the control of ocular accommodation. For some subjects, LCA seems to be the predominant dioptric stimulus for accommodation (e.g., S3 and S6), whereas for others, it is less important (e.g., S4). Fincham 18 found that most (60%) subjects do not accommodate as well in monochromatic sodium light as they do to targets illuminated by white light, and he suggested that the chromatic fringes produced by LCA may play a role in directing accommodation. Figure 7 illustrates how the focus of the eye may be specified by the colored fringes that form on the retina when a point of white light is imaged by the optics of the eye to form the polychromatic point spread function. The point spread functions of Figure 7 were generated for a reduced eye with a 4-mra diameter entrance pupil (r= 5 mm;/= 15 mm; n = 1.333) by varying the standard deviation of a gaussian to simulate different amounts of defocus. 39 The defocus values were based on a formula for chromatic aberration 52 that was applied to calculate defocus for the red (610 nm), green (525 nm), and blue (465 nm) components of the retinal image. Underaccommodation is characterized by a reddish fringe around the point spread function, and overaccommodation is specified by a bluish fringe. Fincham's findings 18 have been corroborated by other investigators, 19 " 21 and the current results support the view that the effects of chromatic aberration (such as the chromatic fringes depicted in Fig. 7) are processed by the visual system for the control of ocular accommodation. It is important to note that the current experiment was conducted using a 3-mm diameter artificial pupil imaged in the eye's entrance pupil. For a pupil larger than 3 mm, the effect of defocus on modulation (contrast) of the retinal image is increased, 40 so the

7 Accommodation to Targets 2701 S2 S hi x pq x fr FIGURE 5. Gain and phase lag of accommodation of two subjects (S2 and S4) who accommodated poorly to long-wavelength stimuli. The illumination conditions from left to right are: 430 nm, 450 nm, 470 nm, 500 nm, 530 nm, 550 nm, 570 nm, 590 nm, 630 nm, 670 nm, NOR, and ACH. Error bars portray 1 SEM response for five trials. D O LJJ CO Q_ effects of LCA on contrast also are expected to be larger. The chromatic fringes in Figure 7 were modeled for a 4-mm pupil, and they would be smaller (in visual angle) for a 3-mm pupil and larger for a 5-mm pupil. Just as the effects of LCA are reduced with a smaller pupil, the magnitude of other aberrations, such as spherical aberration 41 and tranverse chromatic aberration, 4243 are reduced for a 3-mm pupil compared to a 5-mm pupil. Spectral Bandpass Targets The current results suggest that spectrally narrowband targets are poor stimuli for accommodation compared to broad-band targets, for which the chromatic aberration of the eye produces wavelength-dependent differences in focus between the long-, middle-, and short-wavelength components of the white image (see Fig. 7). The magnitude of LCA of the eye is approximately 2 D for targets illuminated by light from the extremes of the visible spectrum. 16 l7 Limiting the spectral bandwidth of a target reduces the effective LCA, and a progressive decline in accommodative accuracy occurs with a progressive decrease in spectral bandwidth For narrow-band targets, such as those used in the current experiment (10-nm bandwidth), the amount of LCA is small (less than 0.1 D). All narrow-band targets in the current experiment were viewed through an achromatizing lens, and the amount of LCA for any particular narrow-band target can be considered to be negligible. Some previous investigators did not find that spectral bandpass targets were poor stimuli for accommodation. 22 " 25 In one of these studies, 23 the experimental paradigm was similar to that used in the current study (sinusoidal target motion in depth). However, the investigators 23 tested only one subject, and it is possible that the subject tested was similar to subject S4 of the current experiment (Fig. 5). Stationary targets that showed no difference between accommodation to white and narrow-band targets were used in two previous experiments, and subjects may have had adequate time to find the position of best focus using voluntary accommodation when the target was stationary. The instruction set used by some experimenters 22 clearly encouraged voluntary accommodation, and volition may have masked the debilitating effect of spectrally band-limited targets compared to broad-band (white) targets. There is some recent evidence to suggest the possibility that the effects of chromatic aberration are less

8 2702 Investigative Ophthalmology & Visual Science, December 1995, Vol. 36, No. 13 S3 S6 CD O LLJ CO Q. FIGURE 6. Subjects S3 and S6 accommodated very poorly to monochromatic targets (crosshatched bars: 430, 450, 470, 500, 530, 550, 570, 590, 630, and 670 nm) compared to the normal white-light target (solid bars). Accommodative tracking ability to a white-light target also was impaired substantially when the longitudinal chromatic aberration of the eye was minimized by an achromatizing lens (singlehatched bars). Error bars portray 1 SSEM response for five trials. important for stationary targets compared to moving targets. 45 However, these data 45 were collected with a target placed at an optical distance of 1 D, which is close to the tonic level of accommodation (approximately 1.5 D; SD = 0.77). 4() Subjects may focus accurately to a target close to 1.5 D, despite a loss of chromatic information from LGA, merely because accommodation decays to this (tonic) level under reduced stimulus conditions. More recent data suggest that when targets are placed away from the tonic level of accommodation (e.g., at 0 D and 5 D), most subjects either lose focus or exhibit greater variability in their accommodative response when the effects of LCA are altered. 47 Previous studies on accommodation to monochromatic targets have used illuminants that are similar to the line spectrum of sodium (590 nm). 18 " 21 One goal of the current investigation was to determine whether the relative loss of dynamic accommodation for 590- nm targets compared to white-light targets 18 " 21 is similar for narrow-band light from other parts of the visible spectrum. The current data suggest that for some subjects (e.g., S3, S4, and S6), targets illuminated by light of 590 nm-wavelength may have been more debilitating to accommodation than mid-spectral targets (500 to 550 nm). Another question that pertains to 590-nm light is whether sodium illumination debilitates accommodation when compared to broad-band white illumination. 48 ' 49 From the current study, it is clear that narrowband illumination does reduce accommodative tracking ability for most subjects, and this suggests that visual tasks that require near vision or changes in fixation at near distances should be conducted under broad-band illumination. In addition, narrow-band visual displays are likely to reduce the ability of the eye to maintain accurate focus, and, where possible, spectrally broad-band displays are recommended. Chromatic Aberration and Color-Opponent Mechanisms Accommodation to spectrally broad-band targets in the presence of ocular LCA is probably mediated by red-green 45 and blue-yellow 44 opponent mechanisms. When a spectrally broad-band target is imaged in the eye, ocular LCA alters the contrast of the target as a function of wavelength When the eye is underaccommodated (hyperopic focus), short-wavelength

9 Accommodation to Targets 2703 JSITY UJ RELA1'IVE II 1ELA1"IVE INTEhJSITY HYPEROPIC FOCUS MYOPIC FOCUS VISUAL ANGLE (arc min) FIGURE 7. Point spread function of the eye for three components of white light (blue = 465 nm; green = 525 nm; red = 610 nm) in the presence of normal chromatic aberration for a reduced eye with a 4-mm pupil. The functions were generated by altering the standard deviation of a gaussian to simulate different amounts of defocus. 39 Underaccommodation (hyperopic focus, 1 D) is characterized by a red fringe, and a blue fringe surrounds the retinal image of a point of white light when the eye is overaccommodated (myopic focus, 1 D). light focuses on the retina, and the contrast of the short-wavelength component of the retinal image is higher than the contrast of the long-wavelength and medium-wavelength components of the polychromatic retinal image. 21 ' 50 Similarly, when the eye is overaccommodated (myopic focus), contrast of the longwavelength component of the retinal image is higher than contrast of the medium- and short-wavelength components. It has been hypothesized that the relative contrasts of the long-, medium-, and short-wavelength components of the retinal image specify the state-offocus of the eye in broad-band illumination It is possible to produce a sine wave grating target in which the relative contrasts of the red, green, and blue components of the retinal image change to simulate defocus in the presence of chromatic aberration. 51 Data from this type of experiment indicate that accommodation does respond to these relative changes in spectral contrast even when the target itself is stationary. 51 This lends support to the hypothesis that spatially bandpass, color-opponent neurons analyze the polychromatic blur spread function of the eye for the control of accommodation. 51 " 53 In light of supporting evidence from previous studies, 51 ' 53 the superior accommodative tracking to a broad-band target with normal LCA in the current experiment was based on analysis by the visual system of relative contrast in the three cone types of the retina, mediated by color-opponent mechanisms. 44 ' 45 Accommodation to Monochromatic Targets Although accommodative tracking in the current experiment was most accurate in white light with normal LCA, some subjects accommodated with moderate gain even in monochromatic light (e.g., SI, S4, and S5). The presence of reasonable tracking ability to monochromatic targets suggests that there may be information other than LCA that provides sensitivity to the vergence (angle of incidence) of light on the retina. 18 In the current study, negative feedback from defocus blur was present, and feedback 54 may have contributed to accommodative tracking of monochromatic targets. However, recent evidence suggests that subjects continue to accommodate to a monochromatic target (550 nm) even in the absence of negative feedback, 55 and this supports the hypothesis that the human eye can determine the angle of incidence of light on the retina for the control of accommodation. 18 Sensitivity to light vergence may vary across subjects, and some subjects may be relatively insensitive to the light vergence in the absence of chromatic aberration (e.g., subjects S3 and S6). With regard to individual variability in accommodation to monochromatic targets from different parts of the spectrum, we speculate that the changes in hue and saturation that occur with changes in angle of incidence of light on the retina (Stiles-Crawford Effect II) 5b57 may be part of a mechanism that specifies the angle of incidence of light reaching the retina. The saturation of monochromatic targets is not constant along the spectral locus, and saturation discrimination thresholds show considerable variability between subjects. 58 Further, the firing rate of color-opponent ganglion cells of the macaque retina to changes in saturation of monochromatic targets depends on the type of spectral opponency of the cell. 59 There is a predominance in the retina of color-opponent ganglion cells with centers that receive input from G cones. 60 ' 61 Saturation discrimination for G + R-retinal ganglion cells is best for monochromatic targets between 470 nm and 540 nm, and neural firing rates decline substantially to changes in purity of spectral targets from the long-wavelength region of the visible spectrum. 59 Perhaps the relative loss of accommodation of most subjects to long-wavelength targets (S2, S3, S4, S6) is related to the relative dominance in the

10 2704 Investigative Ophthalmology & Visual Science, December 1995, Vol. 36, No hl retina of color-opponent cells with G-cone centers. Further research is necessary to identify the stimuli and mechanisms that subjects use to accommodate to monochromatic targets. At spatial frequencies above approximately 5 cyc/deg, asymmetric monochromatic aberrations can have a significant effect on the contrast and spatial phase of the retinal image. 62 ' 63 It is possible that at high spatial frequencies, aberrations that are asymmetric across the pupillary plane 62 ' 63 may form part of a substrate for the control of accommodation. In summary, the current results from monochromatic and white-light targets suggest that broad-band targets with normal chromatic aberration are better stimuli for accommodation than are monochromatic targets (10- nm bandwidth). These results support the view that the longitudinal chromatic aberration of the eye provides a powerful directional stimulus for reflexive accommodation. There is significant intersubject variability in the accuracy of accommodative tracking as a function of peak-wavelength of illumination. When analyzed across subjects, the visual accommodative tracking ability to narrow-band targets is statistically similar for all targets,but most subjects accommodate as well or better in mid-spectral illumination as they do in illumination from the extremes of the visible spectrum. Key Words aberration, accommodation, color, illumination, optics Acknowledgments The authors thank Dr. Steven Mathews for computer programming in Asyst (Keithley Metrabyte), and they thank John Orzuchowski, Matthew Polasky, Diane Schiumo, and Wayne Grofik for technical support. They also thank Dr. Dean Yager and Dr. Angela Brown for consultation on color vision and photometry at an early stage in the design of the current experiment. References 1. Itdeson WH, Ames Jr. A. Accommodation, convergence, and their relation to apparent distance./.p-yychol. 1950;30: Kruger PB, Pola J. Changing size is a stimulus for accommodation. / Opt SocAmA. 1985;75: Judge SJ. Do angular size-change and blur cues interact linearly in the control of human accommodation? Vision Res. 1988;28: Rosenfield M, Gilmartin B. The effect of target proximity on the open-loop accommodative response. Optom Vision Sci. 1990;64: Kruger PB, Pola J. Dioptric and non-dioptric stimuli for accommodation: target size alone and with blur and chromatic aberration. Vision Res. 1987; 27: Kruger PB. The effect of cognitive demand on accommodation. Am J Optom Physiol Opt. 1980; 57: Bullimore MA, Gilmartin B. Tonic accommodation, cognitive demand and ciliary muscle innervation. Am J Optom Physiol Opt. 1987;64: Marg E. An investigation of voluntary as distinguished from reflex accommodation. Am JOptom Arch Am Acad Optom. 1951; 28: Provine RR, Enoch JM. On voluntary ocular accommodation. Percept Psychophys. 1975; 17: McLin Jr LN, Schor CM. Voluntary effort as a stimulus to accommodation and vergence. Invest Ophthalmol Vis Sci. 1988; 29: Bour LJ. The influence of the spatial distribution of a target on die dynamic response and fluctuations of the accommodation of the human eye. Vision Res. 1981;21: Wolfe JM, Owens DA. Is accommodation colorblind? Focusing chromatic contours. Perception. 1981; 10: Raymond JE, Lindblad IM, Leibowitz HW. The effect of contrast on sustained detection. Vision Res. 1984; 24: Owens DA. A comparison of accommodative responsiveness and contrast sensitivity for sinusoidal gratings. Vision Res. 1980;20: Madiews S, Kruger PB. Spatiotemporal transfer function of human accommodation. Vision Res. 1994; 34: Bedford RE, Wyszecki G. Axial chromatic aberration of the human eye. / Opt Soc Am. 1957;47: Howarth PA, Bradley A. The longitudinal chromatic aberration of the human eye, and its correction. Vision Res. 1986; 26: Fincham EF. The accommodation reflex and its stimulus. BrJ Ophthalmol. 1951;35: Kruger PB, Pola J. Stimuli for accommodation: Blur, chromatic aberration and size. Vision Res. 1986; 26: Flitcroft DI, Judge SJ. The effect of stimulus chromaticity on ocular accommodation in the monkey. J Physiol. 1988; 398: Kruger PB, Mathews S, Aggarwala KR, Sanchez N. Chromatic aberration and ocular focus: Fincham revisited. Vision Res. 1993;33: Charman WN, Tucker J. Accommodation and color. J Opt Soc Am. 1978; 68: van der Wildt GJ, Bouman MA, van de Kraats J. The effect of anticipation on the transfer function of the human lens system. Optica Acta. 1974; 21: Lovasik JV, Kergoat H. The effect of optical defocus on the accommodative accuracy for chromatic displays. Ophthalmic Physiol Opt.l988;8: Kergoat H, Lovasik JV. Influence of target color and vergence of light on ocular accommodation during binocular fixation. Curr Eye Res. 1990; 9: Campbell FW, Westheimer G. Dynamics of accommodation responses of the human eye. / Physiol. 1960; 151: Phillips S, Shirachi D, Stark L. Analysis of accommoda-

11 Accommodation to Targets 2705 tive response times using histogram information. Am J Optom Arch Am Acad Optom. 1972;49: Campbell FW. Accommodation reflex. Br Orthoptic J. 1954;11: Westheimer G. The Maxwellian view. Vision Res. 1966;6: Buck SL, Makous W. Calibrating Maxwellian-view optical systems. / Opt Soc Am. 1982; 72: Zhang, Thibos LN, Bradley A. Relation between the chromatic difference of refraction and the chromatic difference of magnification for the reduced eye. Optom Vis Sci. 1991; 68: Kruger PB. Infrared recording retinoscope for monitoring accommodation. Am] Optom Physiol Opt. 1979; 56: Campbell FW, RobsonJG, Westheimer G. Fluctuations of accommodation under steady viewing conditions. J Physiol. 1959; 145: Charman WN, Heron G. Fluctuations in accommodation: A review. Ophthalmic Physiol Opt. 1988;8: Wagner G, Boynton RM. Comparison of four methods of heterochromatic photometry. J Opt Soc Am. 1972;62: Crane HD, Cornsweet TN. Ocular focus stimulator. / Opt Soc Am. 1970; 60: Kurtz M. Handbook of Applied Mathematics for Engineers and Scientists. New York: McGraw-Hill; Winer BJ, Brown DR, Michels KM. In: Chiapetta ME, Vaicunas J, Morriss JM, eds. Statistical Principles in Experimental Design. 3rd ed. New York: McGraw-Hill; Fry GA. Blur of the Retinal Image. Columbus: The Ohio State University Press; Smith G. Ocular defocus, spurious resolution and contrast reversal. Ophthalmic Physiol Opt. 1982;2: Koomen M, Tousey R, Scolnik R. The spherical aberration of the eye. / Opt Soc Am. 1949; 39: Simonet P, Campbell MCW. The optical transverse chromatic aberration on the fovea of the human eye. Vision Res. 1990;30: Thibos LN, Bradley A, Still DL, Zhang, Howarth PA. Theory and measurement of ocular chromatic aberration. Vision Res. 1990; 30: Aggarwala KR, Kruger ES, Mathews S, Kruger PB. Spectral bandwidth and ocular accommodation./opt Soc America A. 1995; 12: KotulakJC, Morse SE, Billock VA. Red-green opponent channel mediation of control of human ocular accommodation. J Physiol. 1995; 482: Liebowitz HW, Owens DA. New evidence for the intermediate position of relaxed accommodation. Doc Ophthalmol. 1978;46: Kruger PB, Mathews S, Aggarwala KR. Accommodation and chromatic aberration: Stationary and moving targets. ARVO Abstracts. Invest Ophthalmol Vis Sci. 1995;36:S Piper HA. The effect of HPS light on performance of a multiple refocus task. Lighting Design Application 1981;February: Corth R. Accommodation and convergence under high pressure sodium illumination. J Illuminating Engineers Soc. 1983; October: Bradley A, Zhang, Thibos L. Failures of isoluminance caused by ocular chromatic aberrations. Appl Opt. 1992; 31: Kruger PB, Mathews S, Aggarwala KR, Yager D, Kruger ES. Accommodation responds to changing contrast of long, middle and short spectral-waveband components of the retinal image. Vision Res. 1995;35: Flitcroft DI. A neural and computational model for the chromatic control of accommodation. Vis Neurosd. 1990;5: Kruger PB, Nowbotsing S, Aggarwala KR, Mathews S. Small amounts of chromatic aberration influence dynamic accommodation. Optom Vision Sci. 1995; 72: Carroll JP. Control theory approach to accommodation and vergence. Am J Optom Physiol Opt. 1982; 59: Kruger PB, Mathews S, Aggarwala KR, Nowbotsing S. Open-loop accommodation: Chromatic and achromatic mechanisms. ARVO Abstracts. Invest Ophthalmol Vis Sci. 1992; 33: Enoch JM, Stiles WS. The colour change of monochromatic light with retinal angle of incidence. Optica Ada. 1961;8: Alpern M, Tamaki R. The saturation of monochromatic light obliquely incident on the retina J Physiol. 1983; 338: Kaiser PK, Comerford JP, Bodinger DM. Saturation of spectral lights./ Opt Soc Am. 1976;66: De Valois RL, Marrocco RT. Single cell analysis of saturation discrimination in the macaque. Vision Res. 1973;13: Ingling CR Jr, Martinez-Uriegas E. Simple-opponent receptive fields are asymmetrical: G-cone centers predominate. J Opt Soc Am. 1983; 73: De Monasterio FM, Gouras P. Functional properties of ganglion cells of the rhesus monkey retina. JPhysiol. 1975;251: Walsh G, Charman WN. Measurement of the axial wavefront aberration of the human eye. Ophthalmic Physiol Opt. 1985;5: Walsh G, Charman WN. The effect of defocus on the contrast and phase of the retinal image of a sinusoidal grating. Ophthalmic Physiol Opt. 1989; 9: