Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
OC ULOMOTOR AND SPAT IAL ORI EN TAT I ON FACTORS .. ~
INTRODUCTION Herschel W. Leibowitz The importance of emphasizing oculomotor and spatial orientation derives in part from the history of research in vision. Because one of the most critical aspects of visual performance is the ability to read, visual research and tests have been directed toward evaluating and cor- recting reading ability. There are literally scores of studies in the literature describing the functional relation between visual acuity, contrast sensitivity, and luminance. However, the demands on the vis- ual system in the military context involve much more than reading or appreciation of contrast. Significant additional factors in the mili- tary environment include spatial orientation, prolonged exposure to low ambient illuminances, and concomitant stimulation of the vestibular system. A consequence of the multidimensional nature of visual demands in military operations is that our present understanding, as well as the tests available for evaluating visual performance, is incomplete. In particular, lowered illuminance results in significant changes in the oculomotor adjustments of accomodation and convergence. These changes, as well as their striking individual differences, have minimal conse- quences during daylight observations but become critical under low- luminance observation conditions. It is essential in studying human performance that our efforts be directed toward the "worst-case" con- ditions which include twilight and night observations, degraded visual environments, and observer motion. Not only is performance degraded at night but illusions, spatial disorientation, and distortions of distance and motion are also more likely. The papers in this section serve to define the problem, to document some encouraging progress that has been made during the past years, and to provide guidelines for future research. The important message under- lying these papers is that vision is multidimensional and that it in- volves different kinds of tasks, particularly recognition and spatial orientation, with different functional relationships to ambient illu- minance. It is also clear that to predict visual performance under demanding conditions, it is also necessary to consider the interactions among the visual, oculomotor, vestibular, and proprioceptive systems. Reference has been mace to the fact that the present report repre- sents the first systematic attempt by the Committee on Vision to survey the night vision literature since the end of World War II. Even a 83
84 cursory comparison of the two literatures reveals the increased scope of the contemporary problem. Meeting this challenge will both increase our understanding of basic mechanisms and, at the same time, assure us that recommendations and procedures for military personnel operating in visually demanding environments can be carried out efficiently and effectively.
NORMAL VARIATIONS OF VISUAL ACCOMMODATION AND BIN=UL~ WRGENCE: , SOME IMPLICATIONS FOR NIGHT VISION - D. Alfred Owens One of the axioms of vision science, as well as everyday experi- ence, is that visual performance generally deteriorates under low illu- mination. Since the earliest psychophysical research, results have been presented documenting the sensory changes that occur with reduced illumination, such as diminished color vision, acuity, and contrast sensitivity. More recently, behavioral research on eye movements has shown that these sensory changes are coupled with variations in the efficiency of oculomotor functions, particularly of accommodation and binocular vergence. Although the sensory processes that guide accommo- dation and vergence are not yet fully understood, both responses depend heavily on variables such as spatial contrast and retinal disparity for accurate and efficient control (e.g., Toates, 1972, 1974~. This infor- mation is gradually lost with reduced illumination, and, as a result, the accuracy and operating range of accommodation and vergence progres- sively decrease. These normal variations of oculomotor behavior con- tribute to a number of problems of night vision, and they may help to explain and predict individual differences that are characteristic of nighttime visual performance. THE RESTING STATE OF THE EYES The resting state of the eyes is a key concept for understanding normal variations of oculomotor behavior. Like the skeletal muscles, the eye muscles exhibit continued tonic activity in the absence of stimulation and active movement (Granit, 1970; Burian and van Noorden, 1974~. Contrary to popular traditional theories, however, the eyes' resting (or tonus) posture does not correspond to optical infinity. Rather, it typically falls at an intermediate distance, which varies widely among individuals with normal vision (Leibowitz and Owens, 1978; Owens 1984~. Although the intermediate resting state has been observed by several investigators during the past century (Cogan, 1937; Morgan, 1946; Schober, 1954; Weber, cited by Cornelius, 1861), it was not gen- erally accepted until recently when measurements of accommodation and vergence under open-loop conditions became available. 85
86 Individual Differences Typical values of the resting positions of accommodation and ver- gence are illustrated in Figure 1. These distributions represent measurements taken in total darknesss, referred to as dark focus and dark vergence. Both sets of data were obtained from the same sample of 60 college students (Owens and Leibowitz, 1980~. These data reveal two important and unexpected variations of the resting state of the eyes. First, both resting states exhibit substantial between-subject varia- bility. All subjects had ostensibly normal vision by conventional cri- teria, yet individual dark focus values ranged from about 25 cm to low hyperopia, and dark vergence values ranged from about 60 cm to infinity. These large individual differences are not picked up by standard clini- cal assessment techniques. The second point, which is not immediately obvious from Figure 1, is that the two resting postures for a given individual are often quite different. Although the mean dark focus and mean dark vergence corres- pond to an intermediate distance (76 and 116 cm, respectively), these mean values are significantly different. Moreover, within-subject comparison of the dark vergence and dark focus values illustrated in Figure 1 showed that they are only weakly correlated (r = 0.32~. These findings indicate that an individual's vergence and refractive state in darkness often differ greatly from those obtained under high illumina- tion, and they imply that knowledge of the resting (dark) state of one system (accommodation or vergence) does not allow prediction of the resting state of the other system. Relation to Clinical Indices While individual differences of the dark focus and dark vergence are not detected by standard clinical measures, there is some rather inconsistent evidence for a relation between the resting states and clinical indices. Two investigations reported evidence that the dark focus is inversely related to ametropia; that is, when far points are equated optically, hyperopes tend to have a nearer dark focus than myopes (Maddock et al., 1981; Epstein et al., 19811. A third investigation reported the opposite relation, i.e., that relative to the far point, the dark focus of myopes tends to be nearer than that of hyperopes (Simonelli, 19831. The basis for this discrepancy is not clear, but one possibility is a difference in the def inition of ametropia. Maddock et al. (1981) and Epstein et al. ( 1981) used standard clinical ret raction techniques to determine their subjects' refractive status, while Simonelli (1983) used measures obtained with a polarized Vernier opto- meter of the subjects' accommodative response for a distant monocular target. Comparisons of the dark vergence to measurements of phoria have also yielded inconsistent results. One study reported that dark vergence measures were signif icantly correlated (r = 0.62) with near phoria but not significantly correlated with distance phoria (Francis
87 _ ~0 Cal , ~ - IX CY - CD ~ Dud n be JO E t' . -or - ~ l 1 1 1 I ~t _ ~O ~_ 8 Q.} _ C) > o o o ~ o ~. . . ~ no ~ flu a nbeJ ~ - 8 - o - - ~lu c Vat --I a 0 .o In s 0 s JJ In · - · In set a, ~5 :' tQ ~ Q 0 u O 0 ~ In u' tQ 0 QJ v ~ · - ~ ~Q y ~ a, CQ v e o a' ~Q u] o ~ · - o ~ o co v Q U] .^ . - . - o ~5 o o U] . - O (q ~ _I O en ~ ~: C~ . o 11 - ~5 o V 3 o a ~4 a a) ~5 U] V ~ O · - ~: · - U] a U] U] · - {Q s" ~ ~ tn u] ~ ~ v s ~ o co - Y N n3 3 o a) Q JJ s ~n a) o .. o U]
88 and Owens, 1983~. These differences are probably related to differences in the stimuli for accommodation and vergence in particular test situa- tions and to complex synergistic interactions of accommodation and ver- gence that arise from stimulation of either system. THE RESTING STATE AND VARIATIONS OF OCULOMOTOR PERFORMANCE The intermediate resting state concept has helped to clarify nor- mal variations of oculomotor performance that occur under degraded stimulus conditions. In general, accommodation and vergence responses tend to be biased toward the subject's characteristic resting posture, and these biases increase progressively as stimulation is reduced. Dark Focus Bias Johnson (1976) was among the first to report progressive biases of accommodation with reduced illumination. He used a laser optometer to measure accommodative responses for a high-contrast target, similar to a reticle, viewed monocularly in Maxwellian view at four luminance levels ranging from 0.051 to 51.4 cd/m . His results for four subjects are presented in Figure 2 as mean accommodative response functions under the four luminance conditions. Under the highest luminances, all sub- jects exhibited accurate focusing performance, with response functions approximating the ideal prediction indicated by the dashed line. As luminance was reduced, however, accommodation became increasingly myopic for stimuli nearer than the dark focus. At the lowest luminance, accom modation remained near the dark focus regardless of stimulus distance. These data show that the operating range of accommodation gradually collapses toward the dark focus, resulting in functional presbyopia, as illumination is reduced through low photopic to mesopic levels. This tendency is also found with more complex natural stimuli. In one study, Leibowitz and Owens (1975a) measured the refractive state of 30 subjects while they viewed total darkness and a distant outdoor scene at three luminances ( full daylight and with light reduced 1. 95 and 4.2 log units by neural-density filters). The results are illustrated in Figure 3 as scatter diagrams comparing individual dark focus values with the sub- jects' accom~r.odative responses for the outdoor scene. In this case, accurate focusing responses would yield data falling along the horizon- tal line at 0 diopter~ (D), and accommmodative responses equivalent to the sub jects' dark focus would yield data falling along the diagonal theoretical line. Consistent with Johnson's (1976) results, as lumi- nance was reduced from daylight to an approximation of nautical twi- light, accommodative responses shifted progressively toward the sub- jects' characteristic dark focus values. As a result, subjects with a relatively near dark focus became increasingly myopic for the distant scene, while these with a relatively distant dark focus showed little change in refraction. It is important to note that biases of accommodation toward the dark focus are not conf ined to conditions of low illumination. Indeed,
89 3.0 J.G. ~ = 1 ~ 2.O Q o _ ~ i.0 D.J .~ ~ ,, ~ it' it' a' -A 30 o 2.0 l.0 ! ~! 1 1.0 2 C) 3.0 0 1.0 2 0 SO QGcominodative St,'m~l~> Distance (Dioptefs) t ~ S K. >=I ~D Am. =I .? 5142 cd,~rr2 ' 6.14 cd/m2 Q5i cdim2 O.~! C~,,~2 I',' in, ~ FIGURE 2 Accommodative response functions of four subjects for a mono- cular reticle stimulus viewed at luminance levels ranging from 51.4 to 0.051 cd/m2. Individual dark focus values are indicated by the black arrows. As luminance decreases, accommodative responses are increas- ingly iased toward the subject' s dark focus, producing anomalous ref. rac- tive errors for targets farther ~ night myopia) and nearer ~ night hyper- opia) than the resting posture. Source: Johnson ( 1976~ .
so 3 2 - DAYLIGHT / r=0.38 / 1 / cn rat _~ Q o ~ _ `_7 . ~ ~ 2 - . _ o o O o £ o c: Cl: 3 2 1 / 7 . . ~. ~ ,( - 's ~ o . i.95 fog units Coffer/ r=0.63 / ~_.~e _-'/ x,,, ~ . .s . a.2 log units town r = 0.70 /~ :,~' /,,' ~ ,, . , - ., i'. , / . . O // 1 2 3 Focus in Tofol [darkness (droplets) FIGURE 3 Scatter diagrams illustrating the relationship between indi- v~dual differences in the dark focus and night myopia. Each diagram compares the sub jects' dark focus with their accommodative responses for a distant outdoor scene at the indicated light level. As luminance decreased from daylight to nautical twilight, focusing responses shifted toward the dark focus, indicated by the diagonal theoretical lines. Thus, subjects with a near dark focus exhibited correspondingly greater twilight and night myopia. Source: Le~bowitz and Owens ( 1975a) .
91 they appear under a wide variety of commonly occurring viewing condi- tions, for example, (1) when viewing a bright empty field (Whiteside, 1952; Westheimer, 1957; Leibowitz and Owens, 1975b); (2) when using optical instruments (Hennessy, 1975) or viewing through small pupils (Hennessy et al., 1976~; (3) when viewing distant objects through an intervening surface or screen positioned near the distance of the sub- ject's dark focus (i.e., the Mandelbaum effect) (Owens, 1979~; and (4) when reading text presented by projection devices (Kintz and Bowker, 1982) or a video display terminal (Murch, 1982~. Dark Vergence Bias Although studied less extensively, analogous biases of vergence toward the subject's dark vergence posture appear when fusional stimr uli are reduced. This was first demonstrated by Ivanoff and Bourdy (1954), who found increasing fixation disparity (i.e., vergence errors) as the luminance of a peripheral binocular stimulus decreased to sco- topic levels. The pattern of fixation disparity varied with the sub- ject and the distance of the target. In general, convergent fixation disparities were found with a distant target, and divergent fixation disparities were found with a near target. The magnitude of these fixation errors varied among subjects, however, with some subjects showing a greater bias toward near fixation (i.e., larger convergent fixation disparity), while others showed a greater bias toward far fixation (lee., larger divergent fixation disparity!. This finding foreshadowed later evidence for individual differences in dark ver- gence (Figure 11. More recently, Francis and Owens (1983) found that changes in vergence accuracy are related to the subject's dark vergence posture. Subjects viewed dim binocular stimuli presented over a range of dis- tances, with retinal eccentricities ranging from 2 to 8 degrees. With increasing retinal eccentricity, vergence responses were progressively biased toward the individual's dark vergence distance, resulting in convergent errors for farther targets and divergent errors for nearer targets. Thus, as fusional stimulation is degraded, the vergence sys- tem is biased toward its resting state in much the same manner as the accommodative system. Summary Research conducted mostly during the past decade has led to a new conception of the behavior of accommodation and vergence. In opposi- tion to traditional views fundamental to modern clinical practice, vergence and accommodation do not relax at the far point of their operating range. Rather, they typically rest at an intermediate dis- tance, which averages about 1 m and varies widely among normal sub- jects. When visual stimulation is degraded, as with reduced illumi- nation, accommodation and vergence responses are progressively biased toward their respective resting postures. These biases produce
92 anomalous refractive errors and fixation disparities that vary as a function of viewing distance in accordance with the individual's char- acteristic resting states. While the dark focus and dark vergence biases are functionally similar, one should keep in mind that they differ in two important res- pects. First, for many individuals the two systems are biased toward different distances. Individual differences in the correspondence (as opposed to the absolute values) of the two resting postures may have important effects on oculomotor performance, particularly under moder- ately degraded viewing conditions. To my knowledge, this possibility has not been investigated systematically. The second distinction is related to differences in the sensory control of accommodation and ver- gence. Since these systems depend to some extent on different types of sensory feedback (e.g., spatial contrast versus retinal disparity), a particular type of stimulus degradation can selectively affect one system more than the other. For example, with reduced illumination, dark focus biases are found at low photopic levels (Leibowitz and Owens, 1975a), while dark vergence biases do not appear until near scotopic levels (Ivanoff and Bourdy, 1954~. On the other hand, dark vergence biases might be obtained more readily with monocular viewing or with binocular stimuli comprised of only horizontal contours, although this has not been specifically tested. The important point is that accommodative and vergence performance may vary in complex ways under naturally occurring conditions. Although the intermediate resting state concept allows simple prediction of oculomotor behavior under severely degraded conditions (e.g., total darkness or a bright empty sky), predictions may not be so straightforward under moderately degraded conditions (e.g., the natural terrain at twilight). I MALI CATI ONS FOR ~ IGHT VI SI ON Both practical experience and findings in the scientific litera- ture show that individuals vary greatly in their visual capabilities at night, and these differences are not fully predicted by standard vision tests. It is likely that a significant portion of these un- predicted differences results from changes in oculomotor behavior described in the preceding sections. This research suggests, for example, that night myopia results largely from the dark focus bias of accommodation and that optical corrections based on the dark focus can ameliorate the effects of such anomalous refractive errors. Other studies suggest that the dark vergence bias is related to problems of visual localization that occur under low illumination. These findings indicate that assessment of the dark focus and dark vergence may pro- vide valuable information for predicting and enhancing nighttime visual performance.
93 Night Myopia For nearly 200 years, night myopia has been recognized as a source of reduced visual detection and resolution capabilities under low illu- mination (Knoll, 1952; Levene, 1965~. Despite repeated efforts, the problem remained intractable primarily because of individual differ- ences in the magnitude of night myopia. While some subjects exhibited as much as 4 D of night myopia, others had no night myopia. It was clear that the same night correction could not benefit everyone, but it was not clear why individuals differed in their susceptibility to night myopia. Evidence for individual differences in the resting state of accommodation suggested a simple solution to this problem,. As illustrated in Figures 2 and 3, accommodation is increasingly biased toward the dark focus when luminance is reduced. The anomalous refractive errors that result from this dark focus bias depend on the individual's dark focus. Therefore, it should be possible to predict and correct anomalous refractive errors, such as night myopia, by simply measuring a person's dark focus and providing a negative opti- cal correction to compensate for the dark focus bias. At least four studies have attempted to eliminate anomalous myopia by this strategy. Preliminary results were quite encouraging, but as will be seen, more work remains to be done. In general, the visual enhancement resulting from these special corrections depends on two factors: (1) the individual's characteris- t~c dark focus, and ~ 2 ~ the quality of available stimulation. Greatest benef its are usually obtained under severely degraded stimulus cond i- tions for subjects who have a relatively near dark focus. Owens and Leibowitz (1976) investigated the utility of optical cor- rections based on the dark focus for vision during night driving. In one experiment, accommodation was measured while the subjects viewed a Landolt ring presented on a white circular field (1 degree diameter) in an otherwise dark visual alley. Luminance was varied from 1.7 to 68.5 cd/m2, and stimuli were viewed binocularly at a distance of 5 m. Unaer all conditions, accommodative responses were highly correlated with the subjects' dark focus values (r = 0.84 to 0.93), indicating that subjects with a near dark focus had correspondingly greater night myopia. The magnitude of this night myopia was not equivalent to the dark focus, however. Instead, accommodative responses were approximately a halfway compromise between accurate focus (i.e., 0.2 D) and the individual's dark focus. This finding suggests that accommodation is still mooer- ately responsive under nighttime driving conditions, which is consis- tent with the progressive nature of the dark focus bias. Subsequent experiments, conducted under laboratory and field con- ditions, compared the effectiveness of three nighttime optical correc- tions: (a) piano; (b) a dark focus (DF) negative correction, which was equal in power to the dark focus, thus placing the dark focus at infin- ity; and (c) a DF/2 correction, which was half as strong as the DF cor- rection. Both laboratory and field data indicated that the DF/2 correc- tion enhanced night vision. Compared with the piano correction, the DF/2 correction improved acuity by as much as 25 percent for sub jects with an average acuity. As expected, none of the corrections made much difference for subjects with a distant dark focus.
94 This finding was confirmed by subjective reports from a field study in which nine subjects were given the three (unlabeled) correction pre- scriptions to use while riding in a car at night. Eight of the subjects selected the DF/2 correction as being noticeably better than either of the other alternatives, and all nine reported that the DF correction was noticeably worse. There was also an interesting exception. One subject, who had an opportunity to test the glasses in fog and snow, reported that although the DF/2 correction was best in clear weather, the DF correction was superior in bad weather. This reaction probably resulted from an increased dark focus bias as visibility deteriorated, and it demonstrates that the same night myopia correction will not be ideal under all viewing conditions. This interpretation is reinforced by two later studies that inves- tigated corrections of empty-field myopia based on the dark focus. Similar to total darkness, the stimulus conditions of these experiments provided no contours or spatial contrast that might serve to stimulate accommodation, and therefore, accommodation would be expected to remain at the subjects' dark focus posture. Post et al. (1979) measured detection thresholds for a small point of light superimposed on a bright uniform background, while the sub- jects wore each of four different optical corrections, i.e., the same three used in the nighttime driving study plus an overcorrection, which was 1. 5 times the power of the DF correction. In this case, the DF cor- rection was best for all subjects. When the detection data were related to nomograms that predict aircraft sighting ranges, the OF corproduced improvements ranging from 26 percent for a subject whose dark focus was 1.0 D to 316 percent for one whose dark focus was 2.0 D. Research on the correction of empty-field myopia was extended by Luria (1980) at the Naval Submarine Research Laboratories in Groton, Connecticut. He measured contrast thresholds for targets ranging in size from 1 to 50 min or arc, with and without optical corrections based on the dark focus. As expected, the results showed that empty- field myopia poses a problem only for detection of relatively small targets (i.e., less than 8 min of arc) and that corrections based on the dark focus enhanced detection by differing amounts for different subjects. One with a relatively far dark focus improved by almost 600 percent. It is interesting to note that the two subjects who exhibited the greatest and least improvement were both clinically emmetropic. Standard clinical refractions indicated that neither subject required corrective lenses. These studies support the hypothesis that anomalous refractive errors, such as night myopia and empty-field myopia, are due primarily to the dark focus bias of accommodation, and they demonstrate that optical corrections based on the dark focus can significantly enhance visual detection and recognition under degraded visibility conditions. In an empty visual field or total darkness' a negative optical correc- tion equivalent in power to the dark focus appears to be optimal. As shown by the nighttime driving study, however, the full DF correction is not appropriate under all conditions, particularly when some (even weak) accommodative stimulation is available.
95 TABLE 1 Percentage of 163 Subjects Who Exhibited Anomalous Myopia of >O. 75 D Under Twilight and Night Luminance Levels Ref ract ion Category Twi l ig fit N ig ht (0.15 cd/m2) (0.001 cd/m2) Myopes 8 33 Emmetropes 24 53 Hyperopes 33 89 . NOTE: These data indicate that hype ropes and emmetropes are more susceptible than myopes to night myopia. SOURCE: Epstein et al. (1981). Since the strength of the dark focus bias varies inversely as a function of stimulus quality, the optimal night myopia correction will vary for different levels of illumination and contrast. This makes the correction of night myopia more complex than originally anticipated. Although the dark focus is of central importance for understanding the mechanism of night myopia and for predicting an individual's suscepti- bility to the problem, it does not provide a universal formula for cor- recting night myopia under all possible conditions. The problem of prescribing night myopia corrections for moderately degraded conditions is illustrated by a Swedish study that used a laser optometer to compare anomalous refractive errors of 163 subjects, who wore their optimal daytime optical correction, at luminances of 120, Ge 15, and 0.001 cd/m2 (Epstein et al., 1981~. Consistent with the results illustrated in Figure 3, they found that the magnitude of night myopia increased from an average of -0.35 D at 0.15 cd/m2 to -1.01 D at 0.001 cd/m2. And they found no systematic correlation between the subjects' normal ret factions and their level of anomalous myopia under the twilight and night luminances. However, they did report the inter- esting finding that corrected hyperopes were most likely (and corrected myopes were least likely) to exhibit significant anomalous myopia. These results, summarized in Table 1, are consistent with the report of Maddock et al. (1981) that the dark focus of hyperopes tends to be more myopic relative to their far point than that of myopes. This suggests that personnel who exhibit the best uncorrected visual acuity under standard test conditions (i.e., hyperopes and emmetropes) are more likely to suffer from anomalous myopia under nighttime operational conditions than are corrected myopes. This reinforces the potential value of knowing an individual's dark focus. Epstein et al. (1981) also tested 29 of their subjects under more natural conditions in which the subject viewed a model landscape
96 containing simulated buildings, roads, cars, and traffic signs under twilight illumination. Using a subjective technique, they determined the negative cor rection necessary for best ecu ity in this situation. All of the sub jects had exhibited 0.75 D or more of anomalous myopia when measured with the laser optometer under similar illumination, but in that case they viewed a monocular pattern of radiating lines. When viewing the model landscape, however, 83 percent of the tested eyes exhibited less than 0.75 D of myopia (12 percent exhibited greater myopia). No simple formula could be used to predict the best correc- tion under the landscape condition from the laser optometer measures, although the optometer measures were useful for identifying subjects who might be susceptible to night myopia. At least three factors may have contributed to the problems of predicting myopia for the twilight model landscape. First, the model landscape may have provided a stronger accommodative stimulus than the fixation stimulus used for the laser optometer measures. If so, the dark focus bias would have been weaker and, therefore, subjects would have experienced less anomalous myopia. Second, binocular vergence may have enhanced accommodation (i.e., vergence accommodation) for the landscape. While monocular viewing was used for the laser optometer measures, Epstein et al. (1981) did not report whether monocular or binocular viewing was used with the landscape. It is well known that binocular vergence enhances the accuracy of accommodation, particularly under low illumination (e.g., Miller, 1980; Kersten and Legge, 1983~. Third, it is possible that variables related to the laser optometer task induced a myopic shift of accommodation that was not obtained in the landscape condition. Post et al. (1985) found that the dark focus of some naive subjects shifts toward a more myopic value when they are asked to judge the speckle motion in a laser test pattern. Summary Clearly, further research will be required to clarify the factors responsible for within-subject variations of anomalous (night) myopia. The available evidence indicates that the dark focus is the best pre- dictor of a subject's susceptibility to night myopia, and it provides a good estimate of the best correction for severely degraded (i.e., sco- topic or empty-field) conditions. For moderately degraded conditions, however, such as a twilight terrestrial scene, accommodation generally strikes a compromise between accurate focus and the individual's dark focus. The refractive error resulting from this compromise will depend on a number of factors, including the subject's dark focus, the quality of available stimulation, and binocular vergence responses (Leibowitz et al., in this volume). Consequently, at present, determination of the best correction for twilight situations will require measurement of accommodation under conditions as similar as possible to the operational situation.
97 Perceptual Illusions Under Low Illumination In addition to limiting visual detection and recognition perfor- mance, oculcmotor adjustments play an important role in localizing objects in space. In particular, accommodation and vergence have been recognized as important sources of distance information since the time of Descartes (Boring, 1942), and an extensive amount of experimental research has documented the influence of these responses on perception of size, distance, and depth (e.g., Lie, 1965; Leibowitz et al., 1972; Foley, 1980~. The recent evidence for biases of accommodation and ver- gence toward intermediate resting postures suggests that these processes may also play a role in illusions of space perception that occur under low illumination. Since the earliest studies of space perception, it has been known that judgments of distance, depth, and size deteriorate with reduced illumination. Such findings have often been cited as evidence for the inadequacy of oculomotor cues for distance. A recurrent finding is the observation that, when distance information is diminished, as in dark- ness, perception appears to be biased toward an intermediate distance. That is, most subjects tend to underestimate the distance of far tar- gets and to overestimate the distance of near targets. They also tend to misperceive size, depth, and motion in a manner consistent with errors of distance perception. These perceptual errors have been studied most extensively by Vogel (1969, 1978), who attributes them to the operation of the "equidistance tendency" and the " specif ic distance tendency. ~' According to this view, as information for distance is re- duced, these autochthonic tendencies cause stimuli to tend to appear in the same depth plane (the equidistance tendency) and at an egocentric distance of about 2 m. Following traditional theories of the oculomotor system, he proposed that these tendencies work in opposition to oculo- motor information for depth and distance. Research on the intermediate resting state of the eyes suggested an alternative hypothesis. Since accommodation and vergence are biased toward an intermediate resting posture under precisely the sort of conditions associated with errors of space perception, it is possible that oculomotor response biases actually contribute to these illusions. This hypothesis was particularly attractive because both the specif ic distance tendency and the oculomotor resting postures exhibit large individual differences. Although this possibility has not been investigated as thoroughly as the relation of the resting state to anomalous refractive errors, the available evidence indicates that oculomotor biases do indeed con- tribute to spatial illusions in low illumination. The first study of this problem simply compared individual differences in the specific distance tendency with the same subjects' dark focus or dark vergence distance (Owens and Leibowitz, 1976~. The results showed that there was no correlation between the dark focus and distance perception in the dark (r = 0.19~; however, there was a significant correlation between the dark vergence distance and the specific distance tendency (r = 0.76~. While correlations must be interpreted cautiously, these results indicate that the dark focus bias has little influence on space
98 perception, and they suggest that the dark vergence bias is related to spatial illusions that occur under low illumination. This possible relation was supported by a subsequent study that investigated the role of oculomotor processes in adaptation to spec- tacles that alter distance perception (Owens and Leibowitz, 1980~. The dark focus and dark vergence postures, as well as perceived distance for a single point of light viewed in the dark, were measured immediate- ly before and after subjects wore experimental spectacles containing negative lenses and base-out prisms. These spectacles forced the sub- ject to accommodate and converge nearer than usual. Consistent with earlier research (e.g., Wallach et al., 1972), subjects reported that these glasses initially caused objects to appear closer and smaller than normal. These effects are thought to reflect the alteration of oculomotor distance information induced by the experimental glasses. After a short time (approx. 10-20 min), however, perception of size and distance return to normal, indicating that some sort of perceptual adaptation to the altered oculomotor information has occurred. If the glasses are then removed, subjects exhibit spatial illusions opposite those originally observed; i.e., objects appear to be farther and larger than normal. This is referred to as the after- effect of adaptation to the spectacles. Comparison of pre- and postadaptation measurements of distance perception and oculomotor resting states indicated that perceptual adaptation may be dependent on adaptive changes in the dark vergence posture. Viewing through the experimental spectacles for 20 min pro- duced a sign recant increase in the perceived distance of the light point after the spectacles were removed (the typical aftereffect), and it resulted in a significant shift of dark vergence posture in the con- vergent direction. There was no change in the dark focus of accommoda- tion, however. These findings show that the resting states of accommo- dation and vergence can be modified selectively, reinforcing the notion that they are determined at least partly by independent mechanisms. In addition, they support the hypothesis that dark vergence specifically plays an important role in space perception, perhaps serving to cali- brate the relation between oculomotor activity and the perception of three-dimensional space (Owens, 1986~. Further research will be necessary to determine the extent to which dark vergence biases that occur under reduced illumination (Ivanoff and Bourdy, 1954; Francis and Owens, 1983) contribute to spatial illusions. The available evidence indicates that such biases affect more than dis- tance perception. A recent study by Post and Leibowitz (1982) found that the dark vergence bias is also related to illusory motion percep- tion. They reported evidence that the magnitude of the vestibuloocular response (VOR) in darkness is determined by the individual's dark ver- gence posture. Consequently, the involuntary compensatory eye move- ~nents tend to induce a slippage of fixation for objects lying nearer or farther than the observer's dark vergence distance. According to Post and Leibowitz (1982), this tendency to lose stable fixation is compensated by activation of the pursuit eye move- ment system. While the supplementary pursuit activity helps to main- ta~n stable fixation, it also causes the fixated object to appear to
99 be moving in the same direction as the pursuit effort. Due to the dark vergence bias, the VOR tends to be too large for objects beyond the resting distance, and the supplementary pursuit effort causes illusory movement in the same direction as the head movement. Conversely, for objects nearer than the dark vergence distance, the VOR tends to be too small, and the supplementary pursuit effort causes illusory motion in the direction opposite of the head movements. For objects lying in the same distance plane as the dark vergence posture, the VOR is appropri- ate, and therefore, supplementary pursuit activity is not necessary and there is no illusory motion during head movements. Summary The evidence indicates that dark vergence biases are related to and perhaps are responsible for misperceptions of distance and motion under low illumination. In view of the close interrelation of perceived dis- tance with other spatial variables, such as size, depth, and velocity, it is possible that the perceptual effects of dark vergence bias are even more pervasive. This possibility is particularly interesting because it suggests that individual differences in spatial localiza- tion abilities at night may be related to the observer's characteristic dark vergence posture. Thus, further research exploring the perceptual consequences associated with the dark vergence bias may help to clarify the basis for individual differences in higher-order perceptual problems that occur under low illumination. RECOMMENDATIONS The resting states of accommodation and vergence may be key factors for predicting and optimizing visual performance under low illumina- tion. Oculomotor response biases toward the dark focus and dark ver- gence can have an important influence on a variety of performance variables, including target detection, recognition, and localization. Moreover, the wide individual differences of the resting states among normal observers imply that an individual's characteristic dark focus and dark vergence values may be valuable predictors of his/her night vision capabilities. Further information about several issues is needed, however, before routine assessment and utilization of resting state measures can be implemented. Assessment Techniques As noted in an earlier section, it is essential to eliminate all stimulation for accommodation and vergence to measure the dark focus and dark vergence postures. This is most easily and commonly accom- plished by making the measurements in total darkness, although a bright Ganzfeld should serve just as well.
100 Dark Focus Most of the research on the dark focus has utilized laser (see Hennessy and Leibowitz, 1972, and Charman, 1974) or polarized Vernier optometers (Simonelli, 1980~. While these instruments have proven to be quite effective for laboratory applications, they may be less suit- able for clinical settings. In particular, recent data show that the subjective task required by these techniques can bias measures of naive subjects (Post et al., 1985~. This implies that an objective technique (in which the subject is passive) would be preferable. Two possibili- ties are suggested: (1) automated (infrared) refractometers, in which the accommodative stimulus can be eliminated; and (2) dark retinoscopy, in which the examiner uses conventional static retinoscopy to measure refraction in an otherwise dark room. It has been found that the ret- inoscope beam is not an adequate stimulus for accommodation, and there- fore the patient's eye remains at the dark focus during the dark retin- oscopy procedure (Owens et al., 19801. Neither of these techniques has been used extensively for assess- ment of the dark focus or night myopia, but in principle, they should both work at least as well as a subjective optometer. .In any case, it is important to remember that the dark focus measure per se may not correspond to the best optical correction for nighttime operations. When prescribing nighttime optical corrections, test conditions should be designed to approximate as closely as possible the level of accord modative stimulation that is likely to be available under operational conditions. Dark Vergence Research on dark vergence has also utilized a subjective measure- ment technique, which is similar to conventional phoria tests, excep t that no accommodative stimulus is provided. One simple version of this approach is to have subjects view a light point that can be positioned at various distances and is flashed for 100 ms at unpredictable inter- vals. A red Maddox rod is positioned in front of one eye so that, on each stimulus presentation, the subject sees a vertical red line with one eye and a white point with the other eye. The subject is asked to report the lateral position of the red line relative to the light point. Over a series of presentations, the distance of the light source is var fed following a bracketing procedure which culminates when the red line and white point appear to be superimposed; at this point, the stimulus source is positioned at the subject's dark vergence dis- tance. It is possible that this procedure affects the subject's dark vergence value (for example, the subject may change vergence when looking for the next stimulus flash), but this has not been investi- gated. It has been found, however, that it is important to remind the subject to relax his/her eyes and not worry if he/she misses some pre- sentations. Also, it seems necessary to allow intervals of about 5 s or more between successive flashes to avoid unwanted hysteresis ef feats.
101 Similar to assessing the dark focus, these potential difficulties might be obviated by a passive, objective recording technique, such as a binocular infrared eye-tracker. In this case, however, the align- ment and calibration requirements of an infrared recording system are probably impractical for clinical screening. Population Norms While it is known that the dark focus and dark vergence postures exhibit large individual differences that are not detected in standard clinical exams, the parameters of these variables in the general (or military) populations are not yet known. Data from college students (Figure 1) indicate that the mean dark focus corresponds to about 1.5 D (67 cm), and the mean dark vergence is about 0.9 m angles (116 cm or 3.2 degrees convergence). To my knowledge there are no data on the dark vergence of subjects outside the college population, but there is some evidence that dark focus values are less "myopic for noncollege subjects (e.~., Epstein et al., 1981~. There is also evidence that, while the dark focus remains intermediate in the accommodative range, it tends to recede with age (Bentivegna et al., 1981; Simonelli, 1983~. Stability and Variation of the Resting State Several studies have shown that individual dark focus values are relatively stable over periods ranging up to 1 year (Mershon and Amerson, 1980; Miller, 1978a; Owens and Higgins, 1983~. Other research indicates, however, that the momentary value of the dark focus can be influenced by a number of situational variables, including near work (Ostberg, 1980; Ebenholtz, 1983; Owens and Wolf, 1983; Schor et al., 1984; Pigion and Miller, 1985), emotional arousal (Westheimer, 1957; Leibowitz, 1976), mood (Miller, 197Bb), and anxiety (Miller and LeBeau, 1982~. It appears that some individuals may be more susceptible than others to such transient changes. For example, Owens and Wolf (1983) found that reading for 1 hour induces a myopic shift of the dark focus in subjects whose initial resting state was <1.5 D, while it induces no change in those whose initial resting state was >3.0 D. In addition, research by Miller and associates indicates that the effects of mood and psychological stress on the dark focus may depend on specific personality traits (Miller, 1978b; Miller and LeBeau, 1982~. To date there are no data on the long-term stability of dark ver- gence for individual subjects. There is evidence, however, that dark vergence is readily modified by visual tasks and by physiological stress. As discussed earlier, it has been found that dark vergence changes with exposure to opposite-base prisms, and these changes may be related to perceptual adaptation to optical displacement (Owens and Leibowitz, 1980; Owens, 1986~. This oculomotor adaptation to pr isms is probably analogous to changes in phoria often observed during clinical auction testing of vergence facility ~ reviewed by Owens and Leibowitz, 1983) .
102 Perhaps of greater concern in the military context, dysfunction of vergence eye movements is the earliest and possibly the most devas- tating visual effect of physiological stress. Early research on vision at high altitudes revealed that hypoxia produces increasing esophoria, convergent fixation disparity, and eventually diplopia (Wilmer and Berens, 1918~. On the ground, such double vision is more commonly experienced as a symptom of illness, injury, and alcohol or drug intoxication (Westheimer, 1963; Wist et al., 1967; Duke-Elder and Wybar, 1973~. Although the relation of these changes to the resting state has not been studied, reports indicate that physiological stressors generally produce a progressive loss of vergence amplitude, with increasing esophoria for distant targets and increasing exophoria for near tar- gets, until finally, binocular fusion is possible only for objects at an intermediate distance. This suggests that stress may induce a dark vergence bias similar to that obtained in darkness. (It is also pos- sible that the stressor influences the dark vergence posture itself.) Whether physiological stress also affects accommodation, and the implications of such effects for visual performance, remain to be investigated. SUMMARY AND CONCLUSI Oli Dur ing the past decade, research on the behavior of accommodation and binocular vergence has yielded new insights, with important impli- cations for night vision performance. It has been learned that when- ever visual stimulation is degraded, as it is at night or in bad wea- ther, both vergence and accommodation tend to shift involuntarily toward the individual's characteristic resting postures, which has been referred to as the dark focus and dark vergence. Although both resting states typically correspond to an intermediate distance, they exhibit wide individual differences, and the two resting postures of - given individual often correspond to different distances. Perceptual consequences of the common tendency of accommodation and vergence to shift toward their respective resting postures also appear to differ. The dark focus bias gives rise to anomalous refractive errors, such as night myopia, empty-field myopia, and the Mandelbaum effect, which can seriously impair visual detection, acuity, and con- t rest sensitivity. These refractive errors are not detected by conven- tional clinical tests, but preliminary studies indicate that they can be predicted and corrected on the basis of the individual's dark focus, with due consideration of the relevant stimulus conditions. Other research has shown that the dark vergence bias is related to misper- ceptions of distance and motion that occur under impoverished stimulus conditions. While further investigations will be necessary to deter- mine the range of perceptual effects resulting from the dark vergence bias, this variable may prove to be a key factor for predicting spatial illusions that are encountered under low illumination. The full implications of variations of accommodation and vergence behavior will not be clear until we understand better the basis of
103 individual differences in the resting states, the range and parameters of their adaptive modification, their role in modulating the synergis- tic interactions of accommodative and vergence control, and the effects of such variables as visual tasks, age, stress, and emotional arousal. The evidence suggests that further research on these problems may yield new and valuable techniques for predicting and enhancing visual perfor- mance in the nighttime environment. REFERENCES Bentivegena, J., D.A. Owens, and K. Messner 1981 Aging, cycloplegia, and the accuracy of accommodation. ARVO abstract. Investigative Ophthalmology and Visual Science (Suppl.) 20:21. Boring, E.G. 1942 Sensation and Perception in the History of Experimental Psychology. New York: Appleton-Century-Crofts. Burian, H.M., and G.K. van Noorden 1974 Binocular Vision and Ocular Motility. St. Louis: C.V. Mosby. Charman, W.N. 1974 On the position of the plane of stationarity in laser refrac- tion. American Journal of Optometry and Physiological Optics . 51:832-838. Cogan, D.G. 1937 Accommodation and the autonomic nervous system. Archives of Ophthalmology 18:739-766. Cornelius, C.S. 1861 Die Theorie des Sehens uno raumlichen Verstellens, pp. 283-285. Halle: H.W. Schmidt. Duke-Elder, S., and K. Wybar 1973 Ocular motility and strabismus. In S. Duke-Elder, ea., System of Ophthalmology, Vol. 11. London: Kimpton. Ebenholtz, S.M. 1983 Accommodative hysteresis: A precursor for induced myopia? Investigative Ophthalmology and Visual Science 24: 513-515. Epstein, D., E. Ingelstam, K. Jansson, and B. Tengroth 19 81 Low-luminance myopia as measured with a laser optomete r. Acta Ophthalmologica 59 :928-943. Foley, J.M. 1980 Binocular distance perception. 87: 411-434 . Francis, E.L., and D.A. Owens 1983 The accuracy of binocular vergence for per ipheral stimuli. Vis ion Re sea rch 23: 13 -19 . Gogel, W .C . 1969 1978 PsYcholog ical Reve iw The sensing of retinal size. Vision Research 9 :1079-1094. Size, distance, and depth perception. In E.C. Carterette and M. Fr iedman, eds., Perceptual Processing. Handbook of Percep- tion, Vol. 9. New York: Academic Press. Granit, R. 1970 The Basis of Motor Control. New York: Academic Press.
104 Hennespy, R.T. 1975 Instrument myopia. Journal of the Optical Society of America 65 :1114-1120. Hennessy, R.T., and H.W. Leibowitz 1972 Laser optometer incorporating the Badal principle. Behavior Research Methods and Instrumentation 4:237-239. Hennessy, R.T., T. Iida, K. Shiina, and H.W. Leibowitz 1976 The effect of pupil size on acco~runodation. Vision Research 16: 587-590. Ivanoff, A., and C. Bourdy 1954 Le comportement de la convergence nocturne ~ The behavior of convergence in night vision ~ . Annals Optique Oculaire 3:70-75. Johnson, C .A. 1976 Effects of luminance and stimulus distance on accommodation and visual resolution. Journal of the Optical Society of America 66 :138-142. - Kersten, D. ~ and G.E. Legge 198 3 Converg ence a cco~runodat ion. America 73 (8~: 332-338. Kintz, R.T., and D.O. Bowker 1 98 2 ~ - - - Journal of the Optical Society of Accommodation response dur ing a prolonged visual search task. Applied Ergonomics 13: 55-59. Knoll, H.A. 1952 A brief history of "nocturnal myopia" and related phenomena. American Journal of Optometry 29:69-81. Leibowitz, H.W. 1976 Visual perception and stress. In G. Borg, ea., Physical Work and Ef fort. New York: Pergamon Press. Leibov?itz, H.W., and D.A. Owens 1975a Night myopia and the intermediate dark focus of accommodation. Journal of the Optical Society of America 65:1125-1128. 1975b Anomalous myopias and the intermediate dark focus of accommo- dation. Science 189:646-648. 1978 New evidence for the intermediate position of relaxed accommo- dation. Documenta Ophthalmologica 46:133-147. Leibowitz, H.W., K. Shiina, and R.T. Hennessy 1972 Oculomtor adjustments and size consistency. Perception and Psychophysics 12:497-500. Levene, J.R. 1965 Nevil Madkelyne, F.R.S., and the discovery of night myopia. Royal Society of London Notes and Reports 20:100-108. Lie, I. 1965 Convergence as a cue to perceived size and distance. Scandinavian Journal of Psychology 6:109-116. Luria, S.~. 1980 Target size and correction for empty-field myopia. Journal of the Optical Society of America 70 :1153-1154. Maddock, ReJ. ~ Me Millodot, S. Leat, and C.A. Johnson 1981 Acco~runodation responses and refractive error. Investioative Ophthalmology and Vi sual Sc fence 20: 387-391.
Murch, G. ~ 982 105 Mershon, D.H., and T.L. Amerson 198 0 Stability of measures of the dark focus of accommodation. Investigative Ophthalmology and Visual Science l9: 217-221. Mille r, R.J . 1978a Temporal stability of the dark focus of accommodation. American Journal of Optometry and Physiological Optics 55 :447-450. 1978b Mood changes and the dark focus of accommodation. Perception and Psychophysics 24: 437-443. 198 0 Ocular vergence- induced acco~runodation and its relation to dark-focus. Perception and Psychophysics 28 :125-132. Miller, R.J., and R.C. LeBeau 1982 Induced stress, situationally-specific trait anxiety, and dark focus. Psychophysiology 19: 260-265. Morgan, M.W . 1946 A new theory for the control of accommodation. Amer ican Journal of Optometry 23: 99-110. How visible is your display? Electr~Optical System Design (March) :43-49. Ostberg, O. 1980 Accommodation and visual fat igue in display work. Pp. 41-52 in E. Grandjean and E. 17igliani, eds., Ergonomic Aspects of Visual Display Terminals. London: Taylor and Francis. Owens, D.A. Owens, D 1976 1 97 5 Th e Mande ~ baum e f f ect: Ev idence f or an accost modat ive b ias toward intermediate viewing distances. Journal of the Optical Society of America 69: 646-652. 1984 The resting state of the eyes. American Scientist 72: 378-387. 1986 Oculomotor information and perception of three-dimensional space. In H. Heuer and A.F. Senders, eds., Perspectives on Perception and Action. Hillsdale, N.J.: Erlbaum. .A ., and H.W . Le ibowitz Night myopia: Cause and a possible basis for amelioration. American Journal of Optometry 53 :709-717. 1980 Accommodation, convergence, and distance perception in low illumination. American Journal of Optometry and Physiolo- g ical Optics 57: 540-550. 198 3 Pe rceptua] and motor consequences of tonic verg ence . In C. Schor and K. Ciufredda, eds., Vergence Eye Movements: Basic and Clinical Aspects. Boston: Butterworth Publishers. Owens , D .A., I . Moh indra , and R. Held 1980 The effectiveness of a retinoscope beam as an accommodative stimulus. Investigative Ophthalmology and Visual Science 19:942-949. Owens, D.A., and K.S. Wolf 1983 Accommodation, binocular vergence, and visual fatigue. Investigative Ophthalmology and Visual Science 19: 942-949. Owens, R.L., and K.E. Higgins 1983 Long-term stability of the dark focus of accommodation. American Journal of Optometry and Physiolog ical Optics 6 0 :3 2-38.
106 Pigion' R.G., and R.J. Miller 1985 Fatigue of accommodation: Changes in accommodation after visual work. American Journal of Optometry and Physiological Optics 62(12):853-863. Post, R.B., C.A. Johnson, and D.A. Owens 1985 Does performance of tasks affect the resting focus of accom- modation? American Journal of OntometrY and Physiological Optics 62(8):533-537. Post, R.B. and H.W. Leibowitz 1982 The effect of convergence on the vestibulo-ocular reflex and implications for perceived movement. Vision Research 22:461:465. Post, R.B., R.L. Owens, D.A. Owens, and H.W. Leibowitz 1979 Correction of empty field myopia on the basis of the dark focus of accommodation. Journal of the Optical Society of America 69:89-92. Schober, H. 1954 fiber die Akkommodationsrohelage. Optik 6:282-290. Schor, C.M., C.A. Johnson, and R.B. Post 1984 Adaptation of tonic accommodation. Ophthalmic and Physiologcal Optics 4:133-137. Simonelli, N.M. 1980 Polarized Vernier optometer. Behavior Research Methods and Instrumentation 12:293-296. l 1983 The dark focus of the human eye and its relation to age and visual defect. Human Factors 25:85-92. . Toates, F.~. 1972 Accommodation function of the human eye. Physiological Review 52:828-863. 1974 Vergence eye movements. Documenta Ophthalmologica 37:153-214. Wallach, H., K.J. Frey, and K.A. Bede 1972 The nature of adaptation in distance perception based on oculomotor cues. Perception and Psychophysics 11:110-116. Westheimer, G. 1957 Accommodation measurements in empty visual fields. Journal of the Optical Society of America 47:714-718. Whiteside, T.C.D. 1952 Accommodation of the human eye in a bright and empty field. Journal of Physiology (London) 118:65. Wilmer, -AH., and C. Ber ens 1918 The ef feet of altitude on ocular functions. Journal of the Amer ican Med ical Soc iety 71 :1394-139 8. Wist, E.R., F.W. Hughes, and R.B. For ney 1967 Ef feet of low blood alcohol level on stereoscopic acuity and f ixation dispar ity. Perceptual and Motor Skills 24:83-87.
ORIENTATION AND LOCALIZATION DURING NIGHT VISION Robert B. Post and Thomas Heckmann INTRODUCTION Spatial orientation is determined by combined visual, vestibular, and proprioceptive activity and the coordination of these activities with motor responses. Under normal, daylight viewing conditions, motion of the observer or changes in orientation produce activity in each of the sensory components. The degree to which a particular sen- sory component is activated depends on the temporal characteristics of the motion or orientation change, as the different systems involved detect different but overlapping temporal frequencies of stimulation (Nasnner, 1970~. A review of the complementary actions of different sensory systems in orientation and their interactions is provided by Dichgans and Brandt (1978~. Visual contours surrounding human observers typically shift oppo- site the direction of self-motion during natural locomotion or vehicu- lar travel. This relative movement of the visual surround provides information for the organism that it is in motion. It also elicits postural adjustments which preserve spatial orientation ana reflexive eye movements which stabilize the retinal images of visual detail. The contribution of vision to spatial orientation may be demonstrated by modifying or eliminating the visual surround or by imposing move- ments of the visual surround on the stationary organism and measuring orientation responses. ORI EN TAT I ON MEASURES Three measures are commonly used to investigate the visual contri- bution to spatial orientation. These are as follows. 1 Postural stabilization responses. Maintenance of stable up . right stance depends on the ability of the organism to detect losses of stability and to compensate by initiating opposing motor activity before stability loss achieves a critical level (a fall). Therefore, the degree to which an individual sways may be used as an index of the adequacy of spatial or ientation. A common technique involves standing 107
108 on a platform which detects forces developed by movement of the body's center of mass. 2. Optokinetic nystagmus. Movement of the visual surround pro- duces compensatory eye movements which minimize retinal image slip of the visual surround. These movements, termed optokinetic nYstagmus, consist of a slow or smooth movement in the direction opposite of head motion (in the same direction of any visual contour motion relative to the observer) alternating with a fast or saccadic phase in the same direction as head motion. The visual control of nystagmus may be demonstrated by moving a visual surround past a stationary observer and measuring the resultant eye movements with any conventional eye- monitoring system. Such visually elicited nystagmus is termed opto- kinetic nystagmus, or OKN. In humans, OKN results from activity in both a cortical pathway, which may also subserve voluntary ocular pur- suit movements, and a subcortical pathway involved in the vestibular control of eye movements (Dichgans, 1977; Baloh et al., 1982~. OKRA resu lting f rom stimulation of the subcortical system continues after removal of the eliciting stimulus, whereas the cortical system shows no such persistence. Continuation of OKh after removal of the visual stimulus is termed optokinetic afternystagmus (OKAN) and provides a convenient technique for determining if the subcortical systems has been activated. 3. Vection. Relative motion of a visual surround which elicits the orientation reflexes of postural stability and OKH also commonly results in sensations of self-motion in the direction opposite that of the surround movement. The induced self-motion sensation is termed vection. Three forms of vection are commonly studied : ( 1) linear vection, the sensation of translators self-motion induced by linearly moving visual contours; (2) circular vection, the sensation of rota- tion about the body' s vertical axis induced by rotations of the visual sur round around this axis; and (3 ~ roll vection, which is induced by sur round rotation around the axis of gaze. Measures of vection fre- quently include latency to the onset of self-motion sensation fol- lowing onset of surround motion or time elapsed until the surround appears stationary (saturation). Although vection refers to the experience of self-motion, roll vection is typically measured by the degree to which the apparent vertical or horizontal appears tilted from the normal orientation. Constraints of Night Viewing As noted above, the visual contribution to orientation responses results from the relative motion of visual contours opposite the direction of self-motion. Under nighttime viewing conditions, ho~v- ever, the visual surround is altered and provides different feeaback during self-motion. First, the surround is typically of lower lumi- nance during night than during day. Because of reduced luminance, much of the potentially available visual detail may be below luminance detection thresholds, thereby reducing the amount of feedback provided or the size of the effective visual field. Another factor stems from
109 situations in which night viewing is typically encountered, such as night flying or driving. During these tasks not only is less detail of the stationary visual surround visible, but contours of the vehicle are also present in the visual field, providing visual detail which is stationary with respect to the vehicle operator. Since these contours are fixed relative to the observer, they oppose the normal role of vision in spatial orientation during self-motion. FINDINGS RELEVANT TO NIGHT VISI ON Postural Stability The contribution of vision to postural stability is demonstrated by the fact that body sway increases when the eyes are closed (Nashner, 1970; Edwards' 1946; Dichgans et al., 1976~. Similarly, if luminance of a visual surround is reduced, there is an attenuation of its effec- tiveness in eliciting postural stabilization responses (Kapteyn et al., 19791. Sway increases with increased distance of the nearest visible detail, in accordance with the reduced relative motion which such con- tours produce during sway (Bles et al., 1980~. Relatively little is known concerning the effects of visual stimulation restricted to dif- ferent locations of the visual field. Bles et al. (1980) have reported that stationary detail in the periphery of the visual field reduces sway while one is viewing distant objects. Although the authors inter- preted this finding as evidence that the retinal periphery dominates orientation responses, data on the effectiveness of central visual field detail on the stabilization of posture was not reported, thereby preventing quantitative comparison of central and peripheral vision. Preliminary measures have been obtained (R.B Post, unpublished data) of postural stabilization by an equivalent area of detail (a light- emitting diode) presented either to the fovea or at an eccentricity of 45 degrees. Postural stabilization was found with the central but not the peripheral stimulus. It has frequently been reported that moving visual contours ais- rupt postural stability in both adults (Dichgans et al., 1972, 1975) and infants (Lee and Aronson, 1974~. Stoffregen (1985) reports that such induced postural sway may depend on the type of optical flow and retinal eccentricity. Specifically, linear sway is induced by either radial or lamellar flow patterns in central vision, whereas only lamel- lar flow induces sway in peripheral vision. Optokinetic Nystagmus OKN may be elicited over a wide range of luminance levels (Gruttner, 1939; Rieken, 1941; Messman, 1944; Mis, 1965), although quantitative analysis of either the gain or the f requency of the response have not been reported. Post ~ in press) has determined that the first phase of OKAN is affected by the luminance of the inducing stimulus, with more luminous stimuli producing more prolonged OKAY.
110 In general, OKN and OKAN are influenced by changes in the size or retinal locus of stimulation. Dichgans et al. {1972) determined that the gain of horizontal OKN increases with the horizontal extent of the stimulus and is largely independent of vertical stimulus extent. A number of studies have indicated that stimulation of the central vis- ual field contributes to a greater degree to OKN than peripheral field stimulation (e.g., Cheng and Outerbridge, 1975; Post et al., 1984~. Optokinetic stimulation of central vision in one direction and simul- taneous stimulation of a larger area of the periphery by contours moving in the opposite direction produces an OKH that is appropriate for the central stimulus (Brands et al., 19731. The direction of OKAN following such stimulation is dependent on the conditions of fixation during the stimulation period. Specifically, OKAN is appropriate for the direction of detail in the central visual field if there was no fixation of stationary detail during the stimulation period and is appropriate for the direction of the peripheral stimulus If fixation of a stationary stimulus was maintained. Vection Linear vection (Berthoz et al., 1975) and circular vection (Leibowitz et al., 1979) can both be elicited by stimuli only slightly more intense than the luminance threshold. Increases in intensity above these levels reportedly have no influence on either the vection onset latency or the duration of vection persistence following stimulus ter- mination. It has f requently been reported that peripheral visual stimulation is more effective than central field stimulation in eliciting vection (for a review, see Dichgans and Brandt, 1978~. Reports of peripheral field superiority include circular vection (Brands et al., 1973), roll vection (Held et al., 1975), and linear vection (Johansson, 19771. In the case of circular vection (Brands et al., 1973), however, com- parisons between a small central field stimulus and a large peripheral field stimulus were emphasized. The results obtained with equal area stimulation of central and peripheral fields apparently reveal no effect of field eccentricity, however. Additionally, roll vection results (Held et al., 1975) obtained with equal area stimulation of central and peripheral fields indicate either small central field superiority, or slight peripheral field superiority depending on the way in which stimulation is restricted to particular eccentricities. Post (1982) obtained no difference in circular vection latency bet- ween equal area stimulation of the central and peripheral fields, and Anderson and Braunstein (1985) report that linear vection is somewhat enhanced with smaller, central displays. Because of such mixed results, it is not possible at this time to conclude whether central or peripheral vision contributes relatively more to vection responses. It is also difficult to interpret results of studies restricting stimulation to particular visual field eccentricities because such stimulation is typically achieved by masking off parts of larger displays. Therefore, not only is moving detail confined to
111 part of the field but the rest of the field contains stationary detail. This methodological factor probably contributes to the common report that increases in stimulus size enhance vection, and small stimuli are incapable of inducing the sensation, since stable visual detail contri- butes to perceived stationarity. VI SUAL LOCALI ZATI of The ability to effectively localize visual detail depends on regis- tration of both the retinal locus of stimulation and the direction of gaze. During conditions of night viewing, either or both of these sources of information may become degraded, thereby disrupting visual localization. With respect to the retinal locus of stimulation, it has f requently been reported that human visual acuity decreases as the luminance level is lowered (see, e.g., Shlaer, 1937; Mandelbaum and Sloan, 1947; Johnson et al., 1981~. There does not appear, however, to be a substantial decrement in the ability to localize single stim- uli. Leibow~tz et al. (1955) report that the radial localization of briefly flashed stimuli is unaffected by wide variation of the lumi- nance of the stimulus above the detection threshold. The independence of perceived location from luminance is perhaps related to a similar independence of the oculomotor system; that is, Steinman (1965) and others (see Boyce, 1967) report that the ability of subjects to fixate the center of a stimulus is not influenced, or is only slightly influ- enced, by variation of the luminance of the stimulus. Whereas the perceived direction of retinal stimulation seems little affected by changes in luminance, it is likely that information about eye position, that is, the direction of gaze, may be severely altered under conditions typical of night viewing, such as reduced visual fields or while viewing scenes containing moving detail. In the absence of any fixation stimulus, subjects apparently have a moderately good sense of eye position, as they can maintain gaze within 1 or 2 degrees of a defined position for prolonged periods (Skavenski and Steinman, 1970; Skavenski, 1971~. The registration of eye position is severely disrupted, however, if the fixated detail is tracked in an otherwise empty visual field. Under these conditions briefly flashed stimuli are mislocalized relative to one another, as if the amount of eye movement is only partially registered (Hansen, 1979; Stoper, 1973~. Similarly, tracked stimuli in an otherwise empty field appear to move more slowly (Fleischl, 1882; Aubert, 1887; Mack and Herman, 1978) and the extent of movement path is underestimated (Mae k and Herman, 1972; Festinger and Easton, 1974~. If multiple moving contours are present, the location of the apparent straight-ahead (ASA) may be displaced by several degrees (Brecher et al., 1972), an effect which can persist or reverse after removal of the moving contours (Post and Heckmann, 1986), causing continued mislocalization in the dark. The illusions described above have most frequently been studied in stationary observers. If the observer is in motion under degraded viewing conditions typical of nighttime, additional distortions of orientation and localization may result. In particular, acceleration
112 causes fixated stimuli in an otherwise empty visual field to appear to move relative to the observer, an illusion termed the oculogYra effect (Whiteside et al., 1965~. A similar effect, termed the elevator illusion, occurs in vertical acceleration during which the plane of the apparent horizontal is displaced and stationary detail appears to move in the same direction as the acceleration (Cohen, 1973) IMPLICATIONS FOR NIGHT AVIATION The preceding discussion has indicated that a number of perceptual distortions occur under viewing conditions similar to those encountered during night aviation. A chief component of most of these effects is that the stable visual environment that is usually present during day- time viewing is either absent at night or replaced by one that moves with the observer. This analysis suggests that one way to eliminate these illusions from the nighttime flying situation is to provide some representation of the stable external environment for the aviator. The recently developed and tested Peripheral Vision Horizon Display (PVHD, 1984), is a device which attempts to do this. It provides an attitude indicator which subtends a large visual angle, and is there- fore more similar to a natural horizon than the small attitude ind~- cators conventionally in use. Use of this or similar devices may help to minimize distortions of or ientation or localization In the night- time flying situation. REFERENCES Anderson, G.J., and M.L. Braunstein 1985 Induced self-motion in central vision. Journal or Experimental Psychology: Human Perception and Performance 11:12~-132. Aubert, H. 1887 Die Bewegungsempfindungen. Pflugers Archiv fur die Gesamte Physiologie 40:459-480. Baloh, R.W., R.D. Yee, and V. Hornrubia 1982 Abnormalities of optokinetic nystagmus. In G . Lennerstrand , D.S. Zee, and E.L. Roller, eds., Functional Basis of Oculo- motor D isorder s . Oxford: Pergamon. B er thoz , A ., ~ . F ava rd , and L . R . You ng 1975 Perception of linear horizontal self-motion induced by per i- pheral vision ~ linearvection): Basic character istics and visual-vestibular interactions. Exper imental Brain Research 23 :4 71-4 89 . B. le s , W ., T . S . K apteyn , I . B. r and t , and F . Ar no ld 1980 The mechanism of physiolo~3ical height vertigo. Acta Otolaryn- gologica 89:534-540. Boyce, P.R. 1967 The effect of change of target field luminance and colour on fixation eye movements. Optica Acta 14:213-217.
113 Brandt, T., J. Dichgans, and E. Koenig 1973 Differential effects of central versus peripheral vision on egocentric and egocentric motion perception. Experimental Brain Research 16:476-491. Brecher, G.A. r M.H. Brecher, G. Kommere~l' F.A. Sauter, and 1972 J. Sellerbeck Relation of optical and labyrinthean orientation. Optica Acta 19:476-481. Cheng, M., and J.S. Outerbridge 1975 Optokinetic nystagmus during selective retinal stimulation. Experimental Brain Research 23:129-139. Cohen, M.M. 1973 Elevator illusion: Influences of otolith organ activity and neck proprioception. Perception and Psychophysics 14:401-406. Dichgans, J. 1977 Optokinetic nystagmus as dependent on the retinal periphery via the vestibular nucleus. Pp. 261-267 in R. Baker and A. Berthoz, eds., Control of Gaze by Brain Stem Neurons. Amsterdam: Elsevier. Dichgans, J., and T. Brandt 1978 Visual-vestibular interaction: Effects on self-motion perception and postural control. Pp. 755-804 in R. Held, H. W. Leibowitz, and H. Teuber, eds., Handbook of Sensory Physiology VIIT: Perception. Berlin: Springer-Verlag. Dichgans, J., T. Brandt, and R. Held 1975 The role of vision in gravitational orientation. Fortsohritte der Zoologie 23:255-263. Dichgans, J., R. Held, L. Young, and I. Brandt 1972 Moving visual scenes influence the apparent direction of gravity. Science 178:1217-1219. Dichgans, J., K. Mauritz, J. Allum, and I. Brandt 1976 Postural sway in normals and atactic patients: Analysis of the stabilizing and destabilizing effects of vision. Agressolog ie 17 :15-24 . Edwards, A.S . 1946 Body sway and vision. Journal of Experimental Psychology 36:526-535. Festroger, L., and A.M. Easton 1974 Inferences about the efference system baseo on a perceptual illusion produced by eye movements. Psychological Review 81:44-58. Fleischl, E. 1882 Physiologish-optisohe Notizen. Sitzungber iohte der Koniglichen Akademie der ~issensohaften in Wien 86:17-25. Gruttner, R. 1939 Experimentelle Untersuchungen uber den optokinetischen Nystagmus. Zeitschrift fur Sinnesphysiologie 68:1-48. Hansen, R.M. 1979 Spatial localization during pursuit eye movement. Vision Research 19:1213-1221.
114 Held, R., J. Dichgans, and J. Bauer 1975 Character istics of moving visual scenes inf luencing spatial orientation. Vision Research 15 :337-365. Johansson, G . 1977 Studies on visual perception of locomotion. Department of Psychology Report No. 206. University of Uppsala, Sweden. Johnson, C.A., J.L. Keltner, and F.G. Balestrery 1981 Static and acuity profile perimetry at various adaptation levels. Documenta Ophthalmologica 50:371-388. Kapteyn, T.S., W. Bles, T. Brandt, and E.R. Wist 1979 Visual stabilization of posture: Effect of light intensity and stroboscopic surround illumination. Agressologie 20:191-192. Lee, D.N., and E. Aronson 1974 Visual proprioceptive control of standing in human infants. Perception and Psychophysics 15:529-532. Leibowitz, H.W., N.A. Meyers, and D.A. Grant 1955 Radial localization of a single stimulus as a function of luminance and duration of exposure. Journal of the Optical Society of America 45:76-78. Leibowitz, H.W., C.L. Shupert-Rodemer, and J. Dichgans 1979 The independence of dynamic spatial orientation from luminance and refractive error. Perception and Psychophysics 25:75-79. Mack, A., and E. Herman 1972 A new illusion: The underestimation of distance during pursuit eye movements. Perception and Psychophysics 12:471-473. 1978 The loss of position constancy during pursuit eye movements. Vision Research 18:55-62. . Mandelbaum, J., and L.L. Sloan 1947 Peripheral visual acuity. American Journal of Ophthalmology 30:581-588. Messman, A.A. 1944 Uber e in Pro jectionsadaptometer zur sub jectiven und ob jectiven Adaptometr ic. Klinische Monatsblatter fur Augenheilkunde 110:446-459. Mis, M. 1965 Den fur optokineticohe Reize empf indliche Netzhautbezirk. Klinische Monatsblatter fur Augenheilkunde 146 :853-862. Nashner, L. 1970 Sensory feedback in human posture control. FIVT 70-3. Cambr idge, Mass.: Man Vehicle Laboratory ~q.I .T. Post, R.B. 1982 Stimulus control of circular vection and optokinetic nystagmus. Ph.D. dissertation, Pennsylvania State University . Induced motion considered as a visually induced oculogyral press illusion. Unpublished paper. Post, R . B ., and T . Heckmann 1986 Induced motion and apparent straight-ahead during prolonged stimulation. Perception and Psychophysics 40: 263-270.
l 115 Post, R.B., C.L. Shupert, and H.W. Leibowitz 1984 Implications of OKH suppression by smooth pursuit for induced motion. Perception and Psychophysics 36:493-498. 1984 Peripheral Vision Horizon D isplay (PVHD) Peripheral Vision Horizon Display. NASA Conference Publica- tion No. 2306. National Aeronautics and Space Administration, Washington, D.C. Rieken, H. 1941 Objektive Adaptometric. Klinische Monatsblatter fur Aucenheilkunde 107:1-11. - Shlaer, S. 1937 The relation between visual acuity and illumination. Journal General Physiology 21:165-188. Skavenski, A.A. 1971 Extraretinal correction and memory for target position. Vision Research 11:743-746. Skavenski, A.A., and R.M. Steinman 1970 Control of eye position in the dark. 10:193-203. Vision Research Steinman, R.M. 1965 Effect of target size, luminance and color on monocular fixation. Journal of the Optical Society of America 55:1158-1165. Stoffregen, T.A. 1985 Flow structure versus retinal location in the optical control of stance. Journal e~ Experimental Psychology: Human Perception and Performance 11:554-565. Stoper, A.E. 1973 Apparent motion of stimuli presented stroboscopically during pursuit movement of the eye. Perception and Psychophysics 13:201-211. Whiteside, T.C.D., A. Graybiel, and J.I. Niven 1965 Visual illusions of movement. Brain: A Journal of Neurology 88: 193-210.
CORRECTION OF NIGHT MYOP IA: THE ROLE OF VERGENCE ACCC` - ODATION Herschel W. Leibowitz, James B. Sheehy, and Kenneth W. Gish In the late eighteenth century, the British astronomer Lord Maskelyne noted that although he did not require any optical correction during daytime observations, he could see more clearly at night with the aid of a negative spherical correction. Because the tendency for many observers to become nearsighted at night, referred to as night myopia, can degrade the detection of small objects and reduce spatial resolution, this anomalous phenomenon has been of special concern to the military. Night and other anomalous myopias have also been of continued theoretical interest since they represent an exception to the normal adaptive behavior of the accommodative system (Leibowitz and Owens, 1978~. In the mid-1950s, Schober (1954) suggested that night myopia is a consequence of the intermediate resting position or dark focus of accom- modation. He pointed out that, contrary to classical theory, a number of visual phenomena could be more parsimoniously interpreted as a nat- ural consequence of the tendency of the eye to return passively to an intermediate focal distance when the efficiency of the accommodative loop is reduced by lowered luminance. More recently it was discovered, in confirmation of the suggestion by Schober, that in addition to the fact that the resting position of the eye is on the average at an inter- mediate distance, there is wide intersubject variability (Leibowitz and Owens, 1975a). This variability permits prediction of the magnitude of night myopia as well as other anomalous myopias, on an individual basis. The general rule is that accommodation tends to return to an individu- ally characteristic resting position when luminance ~ s lowered or under other conditions which interfere with accommodative ef f iciency or when accommodation is no longer effective in improving the quality of the ret inal image . Figure 1 presents the dioptric distance of acconunodation by subjects when they viewed the low-contrast roof of a distant building as a function of their individually characteristic focus in darkness Supported by grants MH 08061 f rom National Institute of Mental Health and EY 03276 f rom the National Eye Institute, and by a grant f rom the U.S. Naval Air Development Center, Warminster, Pennsylvania. 116
117 3 2 Daylight r=0.38 04~ 0 1 2 3 3 2 1 o 1.95 log units lower r=0.63 14 - 0 1 2 3 ~ 4.2 log units lower r=0.70 2 1 o ·/! : _ ~ , _ ·~'~ · .~< ~ . . l ~1 1 1 0 1 2 3 FOCUS IN TOTAL DARKNESS (diopters) FIGURE 1 Dioptric distance of accommodation as a function of individual resting position at three ambient illuminance levels. Source: Leibowitz and Owens (1975b). remain accommodated (Leibowitz and Owens, 1975b). On this plot, accurate accommodation is represented by a horizontal line corresponding to the distance of the fixated object. If no accommodation were in force, the subjects would ~ ~ ' to their own resting positions, which are repre- sented by the diagonal line of unit slope O As luminance is lowered, accommodative distance shifts systematically from the distance of the fixated object toward the individuals' resting positions. It should be noted that the magnitude of this night (more correctly twilight)
118 myopia is related to the individual dark focus values. Subjects with a far dark focus show little if any anomalous myopia, while those with closer resting positions demonstrate increasingly greater amounts. Identification of the mechanism of night myopia suggested a possi- ble basis for amelioration by wearing a special nighttime correction which compensates, on an individual basis, for the tendency of accost modation to return passively to the individual's resting position. Subsequent research indicated that a negative correction equal to approximately one half of the individual's dark resting position increased resolution in a simulated laboratory task and improved clarity and ameliorated symptoms of visual fatigue during prolonged night driving (Owens and Leibowitz, 1976~. CURRENT STUDIES The first of the present series of experiments was designed to develop a simple technique to demonstrate to drivers the valise of special nighttime driving glasses. In the first study, subjects were seated in the driver's seat of an automobile which was parked in an area with no artificial ambient illuminance. They viewed a square wave grating, illuminated by the automobile's headlights, which could be ro- tated by the subject with the aid of a motor so that the bars were just resolvable. Subjects observed the grating with their normal correction and with additional negative corrections equal to their individual dark focus or one half of this value. The data from this study were unexpected because, unlike the pre- vious simulation and field studies, the additional negative correla- tion generally decreased resolution. A series of laboratory studies conf irmed these paradoxical results. Figure 2 presents the contrast sensitivity from one of these studies in which a 16 cycles per degree (c/deg) sine wave grating was presented for 350 ms under the various correction conditions. The subjects viewed binocularly, and a f ixa- tion point subtending 5 min of arc was present at all times. These data indicate that the dark focus correction degrades contrast sen- sitivity. Analysis of the observation conditions for the various studies revealed a consistent difference between the studies in which night myopia was manifested, a negative correction based on the inaividual'- dark focus improved performance, or both and those in which the same correction had no effect or degraded performance. Specifically, per- formance was either unchanged or moderately degraded when viewing was binocular and a stationary fixation point or contours were available but was improved in those studies without such fixation. This obser- vation suggests that the presence of a stationary fixation point or contours might have activated vergence accommodation so that the addi- tional negative lenses produced an overcorrection (Kersten and Legge, 1983~. Evidence for this hypothesis was obtained in two studies. In the first preliminary study, subjects viewed a 16-c/de" grating, he contrast of which had been set to the threshold value determined in a previous study. Such a grating is intermittently visible (Raymond
119 .6 _ .4 E In In o ~3 ~2 .1 o N=5 Viewing Dlstance=7.85m + - 1 No Lenses 1/4 DF 1/2 DF Full DF VIEWING CONDITION FIGURE 2 Contrast sensitivity as a function of optical correction for a briefly exposed high-frequency grating (binocular observation with a fixation point). DF, Dark focus. and Leibowitz, 1985~. The subjects were instructed to indicate by means of a hand-held timer when the grating was resolvable. No fix- ation point was provided. The percentage of time during three 2-n~in trial blocks that the grating was visible for the various corrections is presented in Figure 3. The data indicate that, unlike in the pre- vious study in which a fixation point was available, the negative cor- rections tended to improve visibility if the slight losses in contrast associated with the negative lenses were compensated. In the second more comprehensive study, both contrast sensitivity and accommodative distance were tested under the following conditions in which a fixation point, subtending 4 min of arc, was continuously present: 1. Monocular observation. 2. Binocular observation. 3. Binocular observation with a one-half dark focus correction. 4. Binocular observation with full dark focus correction. Accommodation was measured with a Badal laser optometer. The speckle pattern was superimposed in the visual field after reflection from a moving drum. The subject indicated the direction of the speckle
120 100 r LL 90 m In 5 LL At fir 701 80 ~ ~ /W 50 No Correction 1/2DF OF VIEWING CONDITION FIGURE 3 Percentage of time a high-frequency grating was visible as a function of optical correction (binocular observation without a fix- ation point). pattern, and the accommodation in force was calculated according to a standard procedure (Hennessey and Leibowitz, 1970~. Contrast sensitivity was determined for a 14.5-c/deg sine wave grating using standard electronic techniques, and exposure was for 500 ms (raster flashed on and off) at a mean luminance of 0.1 ftL. The viewing distance was 6.63 m. The subject reported "yes" or "no" as to the presence or absence of thin vertical bars present during the time the raster was flashed on. For a given trial contrast was increased when the previous trial was a miss and decreased when the previous trial was a hit. The threshold was that contrast yielding yes and no responses equally often. Three estimates of threshold were obtained per condit ion (F igu re 4~. The results indicate a monotonic increase in accommodative distance from the monocular to the full dark focus viewing conditions. I t is particularly noteworthy that the accommodative distance is increased for binocular as compared with monocular viewing, so that the addition of the one-half and the full dark focus corrections extended the accom- modative distance beyond the plane of the stimulus. This resulted in an overcorrection which is reflected in the decreased contrast sensi- tivity with the negative lenses in relation to the unaided binocular viewing condit ion.
121 1.5 z 1.0 a: I 5 - .5 ,~ cn ° ~ t~ of LL Y In '_ - .5 Oh C: En -1.0 o -1 e5 Monocular Binocular 1/2 OF FullDF VIEWING CONDITION b Target ~ OF = - LU In At o AL In O ~ F 1 o o 2 3 FIGURE 4 Contrast sensitivity change and accommodation for various observation conditions and optical corrections. The data of Kersten and Legge (1983) and those from the present study suggest that the presence of a clearly visible stationary fix- ation point or contours can reduce or obviate night myopia so that there is no basis for considering special corrections based on the individual's dark focus of accommodation. The most obvious basis for this improvement is vergence-induced accommodation which counteracts the tendency of the eye to return to its intermediate resting posi- tion. However, in those studies in which the stimulus to vergence accomodation was absent or weak, a correction based on the indivia- ual's dark focus effected an improvement. IMPLICATIONS AND RECOMMENDATIONS These studies indicate that a negative correction equal to appro- ximately one half of the individual's dark focus may or may not improve contrast sensitivity during twilight and nighttime observation condi- tionse For subjects observing a stationary fixation point or contours binocularly and tested under transient conditions, additional negative
122 correction is more likely to impair performance.* However, for mono- cular observation in the laboratory, in the absence of a clearly defined stationary fixation point or contours, and while driving an automobile for an extended period of time at night, an additional negative spherical correction appears to be beneficial. These results suggest that data are needed regarding the role of vergence accommodation, particularly in visually demanding situations outside the laboratory. It should also be pointed out that in the previous studies only representative values of luminance level and negative corrections were employed. Systematic data relating lumi- nance level, optical correction, the adequacy of the stimulus to vergence, and the effect of prolonged observation conditions are necessary if the contrast sensitivity of military personnel is to be fully optimized under twilight and nighttime observation conditions. REFERENCES Hennessy, R.T., and H.W. Leibowitz 1970 Subjective measurement of accommodation with laser light. Journal of the Optical Society of America 60~12~:1700-17G1. Kersten, D., and D. Legge 1983 Convergence accommodation. Journal of the Optical Society of America 73:332-338. r eibowitz, H.W., and D.A. Owens 1975a Anomalous myopias and the intermediate dark-focus of accommo dation. Science 189: 646-648. . 1975b N ight myopia and the intermediate dark-focus of accommodation. Journal of the Optical Society of Amer ice 65 :1121-1128. 1978 New evidence for the intermediate position of relaxed account cation. Documenta Ophthalmologica 46~1) :133-147. 8 Luria, S.M. 1980 Target size and correction for empty f ield myopia. Journal of the Optical Society of America 70~9~:1153-1154. Owens, D.A., and H.W. Leibowitz 1976 Night myopia: Cause and possible basis for amelioration. Amer ican Jou rnal of Optomet ry and Phys iolog ical Optics 53 :709-717. Post, R.B., R.L. Owens, D.A. Owens, and lI.W. Leibowitz 1979 The dark-focus of accommodation as a basis for the correction of empty field myopia. Journal of the Optical Society of America 69~11:89-92. l *Previous studies have indicated that a full dark focus correction improves the detectability of small targets in a uniform bright environment. Since such targets provide no stimulus to vergence, the reservations expressed above would not be relevant to the prob- lem of space or empty-fie~d myopia (Luria, 1980; Post et al., 1979~.
123 Raymond, J., and H.W. Leibowitz 1985 Viewing distance and the sustained detection of high spatial frequency of gratings. Vision Research 25:1655-1659. Schober, H.A.W. 1954 Uber die Akkommodationsruhelage. Optik 11:282-290.
NIGHT VISI ON IN RELATION TO OCULOMOTOR FUNCTION Sheldon M. Ebenholtz It is known that certain aspects of oculamotor control may deteri- orate under night vision. Thus, e.g., "Individuals with weak fusional abilities may develop diplopia [i.e.,] an occasional phoria may break into a tropia due to the decrease in fusional stimuli" (Tredici and Miller, 1985, p. 11~. The general implications, however, of night vision for oculomotor function have not been fully explored, with the consequence that the potential deterioration of oculomotor control may go undetected. Several of these implications are developed below. VISUAL PATTERNS r STATIC AND KINETIC, ARE SOURCES OF CONTROL OK' POSITION AND DIRECTION WITHIN THE OCULO>iOTOR SYSTEM Perhaps the most obvious instance of the role of visual stimuli in oculomotor control are of the disparate visual patterns that drive the vergence eye movement control system (Bielschowsky, 1938; Westheimer and Mitchell, 1969; Burian, 1939) and the optical blur that drives accommodation. There is some evidence for a spatial frequency depen- dency in convergence with longer response latencies at higher spatial frequencies (Frisby and Mayhew, 1982), while accommodation (Owens, 1980) exhibits greater accuracy for medium spatial frequencies than either higher or lower values. Furthermore, the contrast of a target pattern plays a critical role in detection by modulating fluctuations in accommodation, with an increase in fluctuations with a decrease in contrast (Raymond et al., 1984a). Thus it is becoming incr eas ing ly clear that the quality of the retinal image plays a critical role in the control of eye movements. Not yet so clear, however, is the pos- sible role of the number of images or contours in the control process. Since image processing occurs at the receptive field level in discrete or digital form, it seems likely that the sheer number of stimulus This research was supported by NIH Research Grant EY03421 from the National Eye Institute. The author is grateful to S.K. Fisher for aid in obtaining the data represented in Figure 2. 124
125 elements might serve the function of a distributed gain control in an eye movement control system. Figure 1 represents a negative feedback model of the accommodation and convergence control system patterned after those of Toates (1970) and Krishman and Stark {1977~. The two subsystems, convergence and accommodation, have exhibited plasticity in their set points, i.e., their respective resting levels, and are there- fore regarded as adaptive control systems (Ebenholtz, 1981, 1983; Ebenholtz and Fisher, 1982; Schor, 19801. The feed-forward elements, labeled tonus control, receive input from the main feedback loops, which are driven by blur and disparity, respectively, and with sus- tained error feedback cause a change in the system parameters that govern the resting level. This shift is adaptive in that the change in extraocular or ciliary muscle tonus serves to reduce the need for further error correction in the feedback loop. Furthermore, this is a symbiotic relationship because the reduced error correction, in turn, reduces further input to the tonus control mechanism (Ebenholtz, in press). Large individual differences exist in the degree of adapt- ability of these systems (Ebenholtz, 1985; Schor, 1980~. EVIDENCE FOR A DI ST RI BUTED INTERNAL GAIN CONTROL IN THE VERGENCE SYSTEM A preliminary test of the hypothesis of a distributed gain control for convergence was carried out by measuring the magnitude of adaptive shift after a period of sustained vergence for targets with varying numb bers of stimulus elements. The smallest elements always were viewed foveally, while additional elements were added peripherally. Thus, foveal versus peripheral control was confounded with sheer element number, contour length, and shape. To charge the adaptive element, three different groups of subjects maintained binocular fixation for periods of 10 to 15 min on targets at a distance of 20 to 25 cm with only foveal (F) targets or during a sep- arate session with both foveal and peripheral (P) targets. The fixa- tion targets, shown at the center of each pattern in Figure 2, ranged in size from 1.7 to 4.3 degrees while the flanking bars projected 5.7 x 26.5 degrees in the lefthand pattern, 3.2 x 23.7 degrees in the middle pattern, and 2.3 x 18.2 degrees in the rig hthand pattern. The displays were the only patterns visible at about 1 cd m~2 and appeared white. They were made of electroluminescent panels and were distributed on an equidistant surface with a radius emanating from a point approximately midway between the eyes. Before and after the fixation period subjects pointed in the open-loop mode, without sight of their hand, to a posi- tion felt to be immediately below a visual test target. A representa- tion of the test condition used in the condition shown in the middle panel of Figure 2 is shown in Figure 3. A second response measure, lateral phoria, also was taken with a Bausch & Lomb Orthorater. The effects of charging the tonus control mechanism shown in Figure 1
126 c) 5 ~ O - Q en C O o I_ r - - . ~ '~ it. T 3 ct ~ o QO O ~ ~ o a) - - L I a) O ~ a) O a' - - - . _ C) In . _ At' ~ ~- . 1 1~ ~ O ~ - ~ 8 . . a) c: ~ 0 ~ a' CJ, ° ~ Q a' ._ · - o ~ O O ~ . V ~ Q O C) ·- ~Q · - o ~1 `: O Q ~ _' · eQ eq E~ ~Q v S~ ta ~q o - C O O C~, ·55 c~) ~ - o O ~ c) ~ C,) t:: C) 0 3 .,, `: 1 ~: _ ~ ~; O ~ · - O . V ~: ~ · - s ~ ~ V H O h v ~n s v U] ~: p. o s ~: · - ~c
o ~ c ~ o - - ~ ~ · o ~ Q 0 ~ _ ._ i_ Q ° - - o in .0 o c x - 1 .0 127 1 1 ' 111 20° s.o 4.0 3.0 2.0 1.0 o 60° ~ 1 _ , _ 1 _ F+P F F+P F+P F , 111 \ 111 11111 \ 111 60~ rr 1 111 / T n: 16 tS 7 t (min): 10 10 15 ~ (cm): 20 20 25 3.0 _ 2.0 _ ' 1 . ~ 1 1 T . FIGURE 2 Patterns used in the sustained fixation period of each con- dition. Below each pattern are represented the change in pointing distance and lateral phoria, respectively. n, number of subjects for each condition; t, period during which subjects maintained convergence d, distance to fixated target. ;
128 ~' ::: ~''W''-'`~ o.o Cal , , ,+_ _1 _ _ ~ , . 1 ' r34.-l ~ CAT LICE__ ~ ~ .1 FIGURE 3 View of apparatus. Target patterns are those represented in the IT iddle panel of F igure 2.
129 should include a shift in the vergence resting level1 toward the near position and, because of the consequent increase in the need to relax convergence to maintain haplopic vision, an increase in pointing dis- tance. Furthermore, if target number increases the system gain, then both measures should be enhanced in the F+P condition relative to con- dition F alone. The manual measures shown in the upper panel of Figure 2 yielded highly significant pre- to postfixation shifts (P<.05) for both F and F+P conditions. Furthermore, in each case the addition of peripheral stimulation significantly increased the shift in pointing distance over that produced by foveal fixation alone. The ratios of induced shift in pointing distance of condition F+P relative to condition F were 1~94:1, 2.06:1, and 2.78:1 for the three stimulus patterns shown from left to right in Figure 2, respectively. There is the suggestion of a slight increase in gain with target complexity when the middle and righthand targets in Figure 2 are compared, but there was relatively little increase with the shift in target eccentricity from 20 to 60 degrees. Thus, the effect probably saturates rapidly within 20 degrees from the fovea. The phoria data showed significant shifts (P<.05) in the eso- phoric direction in condition F (20 degrees eccentricity) and only in condition F+P under the remaining two conditions. Although more sen- sitive measures of resting vergence would be appropriate, the manual data provide at least tentative evidence for the premise that the distribution and number of images constitute an internal gain control mechanism for the vergence system by increasing the number of error signals in the control loop. If this is true then it is reasonable to expect that the quality of these images, i.e., their contrast and spatial frequency, plays a modulating role in eye movement control. An additional implication may be drawn for the difference so~ie- times found when phenomena are compared during darkness and full illu- mination. The data presented above suggest that, when present during the manual pointing test procedure, fusional stimuli provide a strong disparity vergence signal that serves as an optostatic pattern to guide and maintain fixational stability automatically. As a result, in ver- gence adaptation studies, when tests are made under full illumination for comparison with results of a single foveal test target in darkness (Shebilske et al., 1983), there is a diminished sense of change in apparent distance. This outcome follows from the present findings since the increased number of dispar ity-vergence stimuli, under full illumination, lessens the need for voluntary control.2 Thus, a non- cognitive, noninformational account of changes in perception and Since accommodative convergence was present during the phoria mea- surement, the results are not properly equivalent to the convergence resting level. 2Subjects of the present study uniformly reported that the sustained fixation task was much easier to accomplish when the targets contained both peripheral and foveal elements.
130 performance that frequently accompany changes in level of illumination is possible. In addition to the static aspects of stimulation discussed above, optokinetic stimulation has long been appreciated as a source of eye movement control in the service of gaze stability (ter Braak, 1936; Dichgans, 1977), with the capability of driving the eyes around all three axes (Crone, 1975~. Optokinetic patterns are highly functional in normal environments, serving to match eye and image velocity, but they also serve as an updating system for the vestibular ocular res- ponse (VOR). Since the VOR is a feed-forward system, error correction takes place via an external optokinetic loop, a pursuit system loop, or both (Miles and Lisberger, 1981; Ebenholtz, in press). Accordingly, the consequences of poor image quality for the several systems subser- ving gaze stabilization appear to be considerable. Furthermore, an interesting observation is that optokinetic expansion-contraction patterns may have specialized motion-detecting channels (Beverley and Regan, 1980) and also may drive accommodation and convergence (Ittelson and Ames, 1950; Kruger and Pola, 1985~. Although in the latter case the relation between image quality and stimulus adequacy remains to be established, the functional value of kinetic expansion patterns may lie in their capability to signal changes in distance over a range of distances well beyond the several meters usually assigned to accommo- dation and convergence. PRINC IPAL SOURCES OF OCULOMOTOR-ElASED ILLUSI ONS While it may be granted that static and kinetic sources of optical stimulation play a significant role in the control of eye movements, if the oculomotor control system did not contribute to perception, there would be few implications to be drawn for the understanding of visual illusions. In fact, the conscious registration of disjunctive and conjugate eye positions is capable of signaling distance, direc- tion, target orientation (Ebenholtz, in press) and, in general, the three-dimensional spatial path of a moving target. Consequently, the adequacy of registration of the position and trajectory of the eye determines the likelihood of oculomotor-based spatial illusions. In this respect, perhaps the single most important heuristic principle is that reflexive movements do not tend to be registered in consciousness, whereas voluntarily initiated eye movements do lead to an awareness of gaze direction and path. A typical example in the first instance is found when unregistered after nystagmus causes the movement of retinal images to be interpreted, not as to eye movements but to actual move- ment of the surrounding environment. Table 1 lists relatively common modes of eye movement control according to the extent of the volitional component. Given these two categories of eye movement control, illusions of target position and movement path result from one of two sets of condi- tions. These are (1) maintaining fixation in the presence of perturba- tions of a reflexive nature (Post and Leibowitz, 1985; Whiteside et al., 1965y, and (2) maintaining fixation after set points in adaptive
131 TABLE 1 Eye Movement States with Dominant Reflexive and Volitional Components Reflexive Initiation Vestibuloocular and obolith driven Optokinetic and optostatic Vergence Volitional Initiation Fixation and pursuit oculomotor systems have been altered (Ebenholtz, in press). Since vol- untary movements may be associated with ,-efference (Merton, 1972) and since the latter is known to increase muscle spindle gain (~hitteridge, 1959), it seems reasonable to propose that increased sentience of the volitional states results from the competitive action of the reflexive eye movements. It is thus out of the need to balance voluntary against automatic modes of eye movement control that illusions are produced. On the other hand, competing patterns of control provided by stationary backgrounds (Yee et al., 1983; Raymond et al., 1984b) are either not present or are represented with poor image quality during eye movements in darkness. Consequently, muscle spindle gain can be assumed to be lower during darkness than under full illumination conditions because of reduced y-efferent stimulation. Hence, lower sentience or awareness of eye position and eye trajectory (and, hence, of target movement path) is to be expected in darkness. IMPLICATIONS FOR RESEARCH, SELECTION, AND TWINING At dim illumination levels, the contrast between surface edges and backgrounds and between a figure and the ground diminish. At the same time accommodation to targets falls toward the resting level (Johnson, 1976~. The net effect of these processes is to bias the composition of retinal images in night vision in favor of low spatial frequencies, as if one were scanning an environment through low-band-pass filters. Since these images continue to play an important role in the control of oculomotor performance, however, it seems likely that important changes will occur in the efficiency of oculomotor control in night vision. There appears, however, to be a gap in knowledge of the potential inter- actions between a visual system in photopic, mesopic, and scotopic states of dark adaptation and various aspects of oculomotor control, such as pursuit tracking capability, peak optokinetic response, and the ability to maintain foveal fixation when in conflict with vestibular reflexes or peripheral optokinetic stimulation. On this last point, there is some recent evidence to indicate that a high-spatial-frequency, head-fixed target is critical to the suppression of optokinetic nystag- mus (Howard and Ohmi, 1984), but the role of levels of dark adaptation remains to be examined systematically.
132 As more is learned about the relation between the dark-adapted eye and oculomotor control, it may be advantageous to select aircrew on the basis of characteristics of their contrast sensitivity function. If detection and tracking of low-spatial-frequency images is regarded as an important performance characteristic, then screening for high sens- itivity to low-spatial-frequency targets would be appropriate. It is worth emphasizing, perhaps, that in a moving observer, target fixation, i.e., acquisition, and tracking occur frequently under condi- tions in which the VOR has been triggered. Under these conditions, the suppression of VOR by the foveal pursuit system is the indispensable mechanism underlying both target acquisition and localization. It is reasonable to suppose, therefore, that the larger the bandwidth of one's pursuit system frequency-response function (prior to cutoff), the more efficient the pursuit system will be in canceling perturbations emanat- ing from vestibular stimulation. Since individuals differ in their pursuit system cutoff frequency, the possibility of screening for high cutoff frequencies should be considered. Similar considerations should be given to the development of training procedures to enhance the efficiency of pursuit eye movements made in the context of vestibular stimulation (i.e., during and imme- diately after). Such training should take place with gravitoinertial force vectors approaching those encountered in actual flight. Analogous selection and training procedures should also be explored to examine the foveal pursuit system when it is perturbed by competing optokinetic patterns. Oculomotor performance data also are needed to know whether individuals without adaptive control systems (vergence, accommodation, VOR) suffer a higher than normal frequency of asthenopic and motion sickness signs and symptoms. Individuals should be tested for adaptive effects in dark focus, vergence, and VOR. Finally, and most importantly, there is the implication from this analysis of oculomotor function that full advantage should be taken of the current fund of knowledge to maximize night vision performance among aircrew members. To do so would entail as a long-term objective the establishment of a computerized visual function data base to make possible retrospective studies relating visual and oculomotor function to actual performance measures. Such a data base should cover a broad array of visual and oculomotor parameters thought to be relevant to various aspects of flight performance. The following is a sample of such a set of parameters: 1. dark focus, 2. resting vergence, and 3. temporal dark adaptation functions. The following parameters should be measured at mesopic and scotopic levels: 4. peak OKN for a target of known spatial frequency and contrast, 5. pursuit cutoff frequency for a foveal target of known spatial frequency and contrast, contrast sensitivity function,
133 7. pursuit suppression of the VOR, 8. pursuit suppression of OKN, 9. adaptability of the gain of the VOR, and 10. adaptability of the resting focus and resting vergence. REFERENCES Beverley, K.I., and D. Regan 1980 Temporal selectivity of changing-size channels. Journal of the Optical Society of America 70:1375-1377. Bielschowsky, A. 1938 Lectures on motor anomalies. 1. m e physiology of ocular movements. American Journal of Ophthalmology 21:843-854. Burian, H.M. 1939 Fusional movements: Role of peripheral retinal stimuli. Archives of Ophthalmology 21:486-491. Crone, R.A. 1975 Optically induced eye torsion. II. Optostatic and optokinetic cyc~oversion. Albrecht v. Graefes Archives fur klinisches und Experimentelle Ophthalmologie 196:1-7. Dichgans, J. 19?7 Optokinetic nystagmus as dependent on the retinal periphery via the vest~bular nucleus. Pp. 261-267 in R. Baker and A. Berthoz, eds., Control of Gaze by Brain Stem Neurons. Amsterdam: E1sevier. Ebenholtz, S.M. 1981 Hysteresis effects in the vergence control system: Perceptual implications. In D.F. Fisher, R.A. Monty, and J.W. Senders, eds., Eye Movements: Cognition and Visual Perception. Hillsdale, N.J.: Lawrence Erlbaum Associates. 1983 Accommodative hysteresis: A precursor for induced myopia? Investigative Ophthalmology and Visual Science 24:513-515. 1985 Accommodative hysteresis: Relation to resting focus. American Journal of Optometry and Physiological Optics 62:755-762. In Properties of adaptive oculomotor control systems and press perception. Acta Psychological In press. Ebenholtz, S.M., and S.K. Fisher 1982 Distance adaptation depends upon plastic-~y in the oculomotor control system. Perception and Psychophysics 31:551-560. Frisby, J., and J. Mayhew 1982 m e role of spatial frequency tuned channels in vergence control. Vision Research 20:727-732. - Howard, I.P., and M. Ohmi 1984 The efficiency of the central and peripheral retina in driving human optokinetic nystagmus. Vision Research 24:969-976. Ittelson, W.W., and A. Ames, Jr. 1950 Accommodation, convergence and their relation to apparent distance. Journal of Psychology 30:43-62.
134 Johnson, C .A. 1976 Ef f ects of luminance and stimulus d istance on accommodation and visual resolution. Journal of the Optical Society of America 66:138-142. Krishman, V.V., and L. Stark 1977 A heur ist ic model for the human vergence eye movement system. IEEE Transactions on Biomedical Engineering, BME-24, 44-49. P. B ., and J. Pola What drives accommodation? Changing size and blur in oppo sition. Amer ican Journal of Optometry and Physiological Optics (Abstract) 62:7P. Merton, P.A. 1972 How we control the contraction of our muscles. American 226:30-37. Miles, F.A., and S.G. Lisberger 1981 Plasticity in the vestibulo-ocular reflex: A new hypothesis. Annual Review of Neuroscience 4:273-299. - Owens, D.A. 1980 A comparison of accommodative responsiveness and contrast sensitivity for sinusoidal gratings. Vision Research 20:159-167. Post, R.B., and H.W. Leibowitz 1985 A revised analysis of the role of efference in motion pe rception. Perception 14 :6 31-643 . Post, R.B., C.L. Shuperb' and H.W. Leibowitz Implications of OKN suppression by smooth pursu it for induced motion. Perception and Psychophysics 36: 493-498. Raymond, J.E., IBM. Lindblad, and H.W. Leibowitz 1984a The effect of contrast on sustained detection. Vision Research 24 :183-188. Raymond, J.E., K.L. Shapiro, and D. Rose 1984b Optokinetic backgrounds affect perceived velocity during ocular tracking. Perception and Psychophysics 36: 221-224. Schor, C. 1980 Fixation disparity: A steady state error of disparity-induced vergence . American Journal of Optomet ry and Physiological Optics 57: 618-631. Shebilske, W.L., C.M. Karmichl, and D.R. Proffitt 1983 Induced esophoric shifts in eye convergence and illusory distance in reduced and structured viewing conditions. Journal of Exper imental Psychology: Human Perception and Performance 9: 270-277. ter Brask, J.W .G. 1936 Untersuohungen uber optokinetischen Nystagmus. Archives Neer- landaises de Physiolog ie de 1' homme et des Animaux 21: 309-370. (Translation in H. Colleweijn, Oculomotor System of the Rabbit and its Plasticity. New York: Springer-Verlag, 1981. Toates, F.~. 1970 A model of accommodation. Vision Research 10:1069-1076.
135 Tredici, T .J., and R.E . Miller 198S Night vision manual for the flight surgeon. Special Report 85-3, U.S. Air Force School of Aerospace Medicine, Aerospace Medical D ivision {AFSC ~ Brooks Air Force Base, Texas. Westheime r, G ., and D .E . Mitchell 1969 The sensory stimulus for disjunctive eye movements. Vision Research 9: 749-755. Whiteside, T.D.M., A. Graybiel, and J.I. Niven 1965 Visual illusions of movement. Brain 88 :193-210. Whitter idge, D. 1959 The ef feet of stimulation of intrafusal muscle f ibres on sens itivity to stretch of extraocular muscle spindles. Quarterly Journal of Experimental Physiology 44 :385-393. Yee, R. D., S .A. Daniels, D.W. Jones, R.W. Baloh, and V. Honrubia 1983 Effects of an optokinetic background on pursuit eye movements. Investigative Ophthalmology and Visual Science 24 :1115-1122. ..( .
GENERAL DI SCUSSI ON _ HARVEY: This is one for Bob Post. It has to do with flow patterns in movement. Is there a controversy today about the effects of fixa- tion when you're not f ixating in the direction of movement, where the flow pattern originates from? Or as I understand it, the focus of a flow pattern of motion shifts when you look at a direction different from the motion, and that then can give rise to very strange signals of which way your body is actually moving in space. POST: I don't know if there's a big controversy. In a paper by Regan and Beverley, there is a demonstration that the focus of expansion of a flow pattern corresponds to where fixation is rather than the point of collision, which you're headed toward. This contradicts the previous position of Gibson. BEDELL: Dr. Ebenholtz, I understood you to say that there was not an internal signal for reflex eye movements, whereas there was one for voluntary eye movements. An alternative possibility seems to be that there is an internal signal for both kinds of eye movements, but for an involuntary eye movement, say a vestibular-ocular movement, there's also simultaneously a signal of the head movement, which would tend to cancel out the signal of eye movement. Can you distinguish between those two things? EBENHOLT2: I'd like to add a third hypothesis as well. I have been wondering along with others, for a long time, what in the world muscle spindles are doing in ocular muscles? And so it seems to me that I would like to pose a hypothesis that under reflexive control you simply are lacking the gamma efferent stimulation. And as a consequence, gain is low but is probably registered nonetheless. I am sure that some part of the brain has that recorded but that the gain is so low that conscious centers are not aware of reflexive eye movements. So I think that we have then the possibility of several nice hypotheses that now need to be tested. COPENHAGEN: I had a question about plasticity in the system. It almost seems as though we're approaching Heisenberg's uncertainty prin- ciple where you can get changes in the vestibular-ocular reflex in 10 minutes, so by testing it you're automatically changing it. We can see reorganization in the somatosensory cortex in a period of a few minutes. Would you like to comment on that? 136
137 EBENHOLTZ: I have no doubt that, depending on the starting point of the VOR, certain test procedures may indeed alter the very gain that you want to measure. The problem is complicated because the changes in gain, and also changes in phase that have been induced in VOR, decay at differential rates. Now if you imagine that these gain changes are varying from day to day and from time to time, depending upon the tasks that we have been involved in before getting to the laboratory, then you have differential decays of changes in VOR and phase gain back perhaps to some normal level. So that if you really want to measure VOR with some certainty you probably have to take the measurements after a period of isolation to begin with. At least we have a handle on the variance that one finds in these measures. It's in the very plasticity itself. MACLEOD: I'm curious about this plasticity you're discussing. Specifically, I'm wondering if anything is known about whether this adaptation can be situation contingent. For example, when a spectacle wearer adapts to an inappropriate perturbation of his VOR input, can he eventually learn to deal with the situation in such a way that when he takes them off, the system instantly readjusts? It sounds like a trivial question, but I think it has significant implications for how sophisticated an adaptation process is: whether it simply accumulates vestibular and retinal information and makes a blind correlation or whether it makes use of every sort of cue that is available. EBENHOLTZ: I think it is a very important question. People inter- ested in development of simulators--where anyone in the simulator will make certain head movements at the time when his eye is on the target-- are also very interested in whether these things transfer. Now I know that you can induce changes simultaneously in the VOR and in different directions. For example, from the VOR, changes that I've investigated in my laboratory, one obtains the illusory movement when the head moves in the horizontal direction. It is a very specific adaptive shift. If you move your head up and down (nodding) you do not get illusory move- ment unless you adapted initially also in the vertical plane. So we know that the nature of the plasticity, especially in the VOR, is very highly specific. Can you control the changes in VOR, or perhaps other ocular-motor systems, through cognitive links? There are one or two studies in the literature of prism adaptation that seem to suggest that simply the prism frames were enough to trigger the effect, but I remain rather skeptical. I think rather what is triggering it are head movements. I think you have to get into the system with a particular phase and frequency of head movement before you can trigger it. But it is cer- tainly an open question as to whether we can train multiple adaptations simultaneously and simply trigger them by using the top-gown paradigm. We need some more research on this question. JOHNSON: I have a question for either Hersch or Jim [Sheehy] about trying to come up with rapid, simple, cark-focus corrections for low- luminance situations. I '11 just play devil's advocate here for a second. Why worry about the dark focus at all? Why worry about the theoretical implications? Why not just go out and define a performance task, fit the observer with a series of lenses, and find the corrections
138 for which performance is optimal? You circumvent all the problems of trying to make measurements of oculomotor responses, developing instru- mentation, and then translating the results into something that again gets back to performance. LEIBOWITZ: That is what we were trying to do with the field studies. We did not use any instruments, but rather had subjects sit in the car and rotate the square-wave grating (by means of a motor) illuminated by the automobile headlights until it could be resolved. What was surprising was that in almost every case a correction ba see on the dark focus of accommodation degraded acuity, which, in turn, led us to question our night myopia correction procedure. In practice, there is another consideration because during driving, the problem of seeing small stimuli, which are most affected by refractive errors, is not very great. Although contrast is certainly critical during night driving, there are very few stimuli subtending small visual angles that are cri- tical during nighttime driving. The implication of these data is to question the basis for night vision corrections based simply on the dark focus. Regarding your suggestion, one could follow the procedure practiced by some people when they buy glasses. They just go into a store and try them on. I am confident we can do better than this, but in any event, as you suggest, the final criterion must be performance in the relevant real-world situation. KINNEY: I had another question along the same lines. Did I under- stand you to say that the test in the field was 16 cycles/degree? And did you use any ether one? LEIBOW~TZ: The target, when oriented vertically, presented a 15-cycle/degree square wave. The procedure was to start at a random set point and rotate until the observer could just discern the stripes. he then measured the angle subtending by the target and calculated the threshold in cycles per degree. Contrast of the target, which was i~lu- minated by standard automobile headlights, was 77 percent. KINNEY: Have you tried lower spatial frequencies? I would think that woula be more applicable to the laboratory studies, that this would be a major difference between the two, if the dark focus does not work so well for the higher ones. SHEEHY: Bell, in the laboratory we tried to number different spa- tial frequencies. The f irst one we started with was a 4-cycle/degree pattern, and we found absolutely no effect except that people were less sensitive. So with that we switched to ~ 16- and then 24-cycle/degree pattern, both sinusoids. Again, they were adjusting the contrast of that pattern while wearing their corrections. The results were iden- tical. MONACO: Fred [Owens!, Her sch had mentioned something about the dark retinoscopy. I was wondering whether you might want to elaborate a little on that in terms of its being an effective tool for defining the dark focus value. Is there ongoing work with that? OWENS: To my knowledge, there are a couple of optometrists in pri- vate practice who use dark retinoscopy to prescribe for night myopia corrections for those patients who complain. Dark retinoscopy is just standard static retinoscopy without a fixation target. Now, theory has always held that it is essential to control accommodation; the standard
139 techniques are either by cycloplegia or by having the patient fixate a distant acuity chart during the retinoscopic exam. We were surprised that a retinoscope beam is not an effective stimulus for accommodation, that when a patient is fixating the retinoscope beam without any other visible contours, their accommodation returns to the individual's dark- focus state. So it is possible to measure with a clinical instrument that's readily available. That's the advantage of dark retinoscopy. I don't think we have sufficient data to indicate how effective this technique would be for wide-scale screening. I'd like to take the opportunity to come back to a point that Dr. Johnson raised a few moments ago about "why measure the resting state anyway?" There may be a couple of good reasons. One is that for individuals who recognize their problems with night vision, we found measurement of their dark focus to be quite effective in coming up with a correction that helps them--and without going out in the field and doing a sort of empirical "trial-by-error" determination. Perhaps a better argument, though, is the general importance of ocu- lomotor bonus for visual performance under a wide range of conditions. We have seen evidence that the dark focus predicts the best correction for empty-field myopia, and it is probably the best predictor for very dark conditions. Other data indicate that visual problems associated with the use of optical instruments and with fatigue from near work depend on the individual's resting state. For example, individuals who have a near-dark focus and near-aark vergence are less susceptible to fatigue from near work. I suspect this is related to the sort of plas- t~city that Dr. Ebenholtz was talking about. Symptoms of visual fatigue appear to be related to changes of oculomotor tonus of the kind that Dr. Ebenholtz discussed; the problems might be avoided by setting the working distance to correspond to the individual's resting posture. My point is that we may be able to predict and correct a number of different problems on the basis of the resting state, which would be preferable to separate attempts to solve each problem empirically. The night-driving situation, in particular, may be problematic--because it involves a number of interacting factors, such as variations in low- level response gain and vergence-accommodation synergy.
