Chapter 17 Evaluation of Visual Function CHRIS A. JOHNSON Table Of Contents |
VISUAL ACUITY CONTRAST SENSITIVITY VISUAL FIELD SENSITIVITY CONCLUSION ACKNOWLEDGMENT REFERENCES |
The stimulus for vision is light, which constitutes only a small portion
of the electromagnetic spectrum for wavelengths between approximately 400 and 700 nm. Light
is emitted by natural and artificial sources and
is reflected by objects in our environment. The simplest eyes in nature
function as basic light detectors, whereas the human visual system
has evolved into an extremely complex mechanism for processing visual
information. When light enters the eye, it is refracted by the optical
media to form an inverted image on the retina. The photoreceptors convert
this light energy into electrical signals that are processed by
neural elements in the retina, optic nerve and higher visual centers
of the brain. This processing of visual information is carried out by
various subpopulations of neural mechanisms that are responsible for encoding
stimulus features such as motion, form, color, depth, and other
fundamental properties of the visual image.1–6 The relationship between the physical properties of light and the consequent
perceptual and behavioral responses of the observer is known as
visual psychophysics, which serves as the primary foundation for the
clinical assessment of visual function. Excellent reviews of color vision and binocular vision are available elsewhere in these volumes. The present chapter will therefore be limited to a discussion of three other aspects of visual function that are important parts of a clinical ophthalmic evaluation: visual acuity, contrast sensitivity, and measurement of visual field sensitivity (increment threshold sensitivity) by means of perimetry. These are all visual functions that have a long history of psychophysical and physiologic investigation devoted to them, but they also form an important part of the evaluation of the functional capabilities of patients with ocular disorders. Early detection of eye disease, differential diagnosis of various ocular and neurologic disorders, and monitoring the efficacy of therapeutic regimens are just several of the roles these visual function tests assume in a clinical setting. This chapter provides a brief overview of visual acuity, contrast sensitivity, and visual field testing and their application to the clinical evaluation of patients. |
VISUAL ACUITY | |
Visual acuity is the most common measurement of visual function that is
performed in a clinical setting. It serves as the primary means of assessing
the integrity of the optics of the eye and of the neural mechanisms
subserving the foveal region, because good visual acuity requires
that both optical and neural components are functioning properly.7 Visual acuity is used to monitor the status of central visual function
in patients and is an essential part of the procedure used to prescribe
spectacle and contact lens corrections. In addition, visual acuity
is important for many “real world” tasks, such as reading,8,9 face recognition,10 and other activities involving the discrimination of fine spatial detail. Visual acuity is specified in terms of the visual angle subtended by fine spatial detail. An object's visual angle is determined by its physical size and its distance from the observer. For small angles, the visual angle of an object can be approximated by arc tan(object size/object distance). There are generally considered to be three types of visual acuity measurements: detection acuity, resolution acuity, and identification acuity. Detection acuity refers to the smallest stimulus object or pattern of elements that can be distinguished from a uniform field. The minimum angle of detection (MAD) is primarily limited by the contrast of the stimulus. For normal human foveal vision, Thibos and Bradley11 have recently demonstrated that optical components impose the primary limiting factor on detection acuity by means of contrast attenuation for small stimuli. Resolution acuity refers to the smallest amount of spatial detail that can be discerned to allow one pattern to be distinguished from another. Typically, this type of acuity is measured with the use of a grating pattern of alternating light and dark stripes, and the observer's task is to determine the orientation (horizontal or vertical) of the grating. Resolution acuity is normally specified in terms of the minimum angle of resolution (MAR), and it is also limited primarily by contrast. For normal human foveal vision, resolution acuity is mainly constrained by the properties of the photoreceptor mosaic and underlying neural mechanisms.11 Identification acuity refers to the smallest spatial detail that can be resolved to allow one to recognize objects such as letters. It is this type of acuity measurement that is obtained in a clinical ophthalmic setting. Specification of identification or recognition acuity is usually in terms of the MAR, or values such as Snellen notation, which is based on MAR. For the letters typically used in eye charts, the MAR would correspond to the angle subtended by the thickness or stroke of a letter. The letters for most visual acuity charts are designed such that the overall size of the letter is five times larger than its thickness. An example of a typical eye chart used for clinical evaluation of visual acuity is shown in Figure 1. This design is basically the same as that originally introduced by Snellen in 1862,12 and the most common form of reporting visual acuity is still in terms of “Snellen notation.” This value consists of a fraction in which the numerator is the testing distance (usually 20 ft or 6 m) and the denominator is the distance at which a “normal” observer would be able to read the letter. Visual acuity charts use a reference of 1 minute of arc visual angle for the smallest resolvable detail (MAR) for a “normal” observer, which corresponds to a Snellen fraction of 20/20 (or 6/6). Similarly, a 20/40 (6/12) letter would be 10 minutes of arc high and have a thickness of 2 minutes of arc. It should be noted that the 20/20 reference for “normal” vision was developed more than 100 years ago. With today's high-contrast eye charts and better lighting sources, most normal individuals under the age of 50 years can be corrected to better than 20/20 visual acuity.
A new design of visual acuity charts has recently emerged.13,14 An example of one of these new eye charts (the Bailey-Lovie chart) is presented in Figure 2. This improved design has several advantages. First, the letters used are of approximately equal detectability, as compared with earlier charts that had some letters that were more legible than others. Second, each line has an equal number of letters, as compared with earlier charts with only one or two letters for the 20/100 and 20/200 visual acuity lines, and a large number of letters for the 20/20 and smaller visual acuity lines. Third, the spacing between letters is proportional to the letter size, as compared with an unequal spacing between letters for the older acuity charts. Finally, the change in visual acuity from one line to another is in equal logarithmic steps, as compared with very small changes for different lines at the small-letter end of the chart and rather large changes for the big-letter end of the chart for older eye charts. This new eye chart thus permits the ability to better specify visual acuity in terms of Snellen notation, MAR, or log MAR values. In addition, new methods of scoring responses to this type of visual acuity chart can provide greater sensitivity and reliability of the measure.15,16 The Early Treatment Diabetic Retinopathy Study (ETDRS) visual acuity chart, which is based on the design of the Bailey-Lovie eye chart, has now been used in several multicenter clinical trials sponsored by the National Eye Institute.17–19 A number of stimulus parameters are known to affect visual acuity measurements in normal individuals, including background adaptation luminance,20–22 contrast,22,23 refractive error,24–26 pupil size,27,28 retinal eccentricity,29–31 duration of presentation,32–34 the type of optotype used,35–37 and interaction effects from adjacent visual contours, or “crowding.”38–40 In clinical patient populations, factors such as “crowding,” luminance level, contrast, and type of optotype used can sometimes have an even greater influence on visual acuity measurements than they can for normal individuals, depending on the particular ocular disorder. If test parameters such as background luminance are varied in a consistent manner, they can be potentially useful for diagnostic purposes. However, the unintentional variation of background luminance or other test parameters (from one examination room to another or from one visit to another) can be problematic for monitoring the visual status of patients. The measurement of visual acuity in special populations (e.g., young children and handicapped individuals) is not always possible with a letter chart. For these populations, “preferential looking” techniques, oculomotor responses such as optokinetic nystagmus, and electrophysiologic measures such as the visual evoked potential can be used to estimate visual acuity.41–44 In addition, a number of new eye charts and behavioral test procedures have been developed to assess visual acuity in nonverbal or handicapped individuals.45 These tests use patterned stimuli and caricatures of faces or common objects with “critical detail” to obtain an indication of the individual's level of visual acuity. Haegerstrom-Portnoy has provided an excellent review of many of these new behavioral procedures.45 |
CONTRAST SENSITIVITY | |||||
Although visual acuity is able to specify the smallest spatial detail that
can be resolved for high-contrast stimuli, it does not provide any
information about the visual system's ability to detect and discriminate
objects of different sizes and contrasts. The most commonly used
stimuli for assessing the contrast sensitivity characteristics of
the human visual system are sinusoidal gratings, consisting of an alternating
pattern of light and dark bars whose luminance varies sinusoidally
in the direction perpendicular to the grating's orientation. A
graphical representation of the luminance profile of a sinusoidal grating
is presented in Figure 3.
The appearance of a sinusoidal grating is an alternating series of fuzzy light and dark lines. The size of the light and dark bars of the grating is specified according to spatial frequency, which is the number of cycles (pairs of light and dark bars) of the grating pattern per degree of visual angle. Typically, the range of spatial frequencies examined is between 0.5 and 30 cycles per degree. The contrast of the sinusoidal grating is determined by the luminance of the peaks (Lmax) and troughs (Lmin) of the luminance profile, according to the following equation: Contrast can thus vary from a minimum of 0 for a uniform field (i.e., Lmax and Lmin are equal) to a maximum of 1 (where Lmin is equal to zero). For a particular spatial frequency, the contrast threshold is the minimum amount of contrast needed to detect the presence of the grating pattern. Contrast response properties are typically reported in terms of contrast sensitivity, which is the reciprocal of the contrast threshold (sensitivity = 1/threshold). The human visual system varies in the amount of contrast needed to detect a grating pattern for different spatial frequencies, or sizes of light and dark bars. This can be observed in Figure 4, which presents a sinusoidal distribution of light and dark bars in which spatial frequency varies logarithmically along the abscissa and contrast varies logarithmically along the ordinate. At a standard reading distance, it should be observed that intermediate spatial frequencies (medium-sized bars) are visible at lower contrast levels (i.e., the light and dark bars extend farther vertically) than are low spatial frequencies (wide bars) and high spatial frequencies (narrow bars), giving an inverted U-shaped appearance to the pattern. The use of sinusoidal gratings to evaluate spatial contrast sensitivity properties of the human visual system has several advantages. First, optical defocus does not change the shape or appearance of the sinusoidal grating, other than to reduce its contrast. Thus, only one attribute of the stimulus is altered by blur. Second, the contrast sensitivity function provides a means of characterizing the overall response properties of the visual system to a wide variety of complex stimuli and visual images. In the past, the contrast sensitivity function for vision has been compared to the audiogram for characterizing hearing ability.46 The audiogram specifies the minimum amount of sound energy necessary to detect pure tones (temporal sinusoidal waveforms) for various frequencies of pitch. This information can then be used to determine an individual's ability to hear complex sounds that consist of compound waveforms. In a similar fashion, the spatial contrast sensitivity function for vision presumably makes it possible to determine an individual's ability to process spatial information from complex visual scenes. Fourier analysis demonstrates that any complex waveform can be decomposed into a series of sine waves of various frequencies, amplitudes, and phase (position) relationships.47 Thus, any complex visual scene can theoretically be broken down into a specific combination of sinusoidal distributions of light of different spatial frequencies, amplitudes, and phases in the vertical and horizontal dimensions. If the contrast response characteristics of the human visual system to various spatial frequencies are known, then it should be possible to determine how complex visual scenes will be processed by the visual system. Note, however, that the contrast sensitivity function describes the behavior of the visual system at threshold contrast levels, which is not necessarily representative of the visual system's characteristics for processing suprathreshold contrast information.48–50 An illustration of a normal contrast sensitivity function for the human visual system is presented in Figure 5. Contrast sensitivity is greatest for a spatial frequency of about four to five cycles per degree, with progressive reductions in contrast sensitivity for higher and lower spatial frequencies. This inverted U-shaped function reflects the broad bandpass characteristic of the human visual system's contrast response properties. This shape has been reported to represent an envelope of the response characteristics of multiple quasi-independent channels or neural mechanisms with narrower bandpass response characteristics,51 as illustrated in Figure 6. Indeed, neurophysiologic studies support the notion that the visual system uses a form of spatial frequency analysis in the early processing of visual information.52–54
The use of sinusoidal gratings to study human visual function was first introduced by Schade in 195655 and was followed by a series of important studies by several investigators that established the theoretic and empirical basis for the contrast sensitivity function of the normal human visual system.56–59 These investigations were instrumental in defining the optical and neural components underlying the normal contrast sensitivity function. A number of factors affect the measurement of the normal contrast sensitivity function, including background adaptation luminance,59–61 stimulus field size,62–64 retinal eccentricity,65–68 pupil size,28,56 temporal characteristics,60,69,70 stimulus orientation,71,72 and various optical factors such as defocus, dioptric blur, diffusive blur, and astigmatism.11,37,56,73,74 Clinical applications of contrast sensitivity testing have been quite extensive. A number of investigators have used the contrast sensitivity function to evaluate optical properties of the human eye, including refractive error and defocus,11,37,56,73,74 corneal diseases,75–83 cataracts,84–90 refractive surgery,91,92 intraocular lenses,93–95 and normal aging properties of the optics of the eye.87,88,96–98 In many instances, contrast sensitivity deficits have been reported for patients with normal visual acuity and optical conditions that produce subtle changes in the quality of their vision. A number of retinal disorders have also been evaluated with the contrast sensitivity function, including diabetic retinopathy,99–101 age-related macular degeneration,102–104 and a variety of other retinal disorders.105–107 Several investigators have also reported foveal contrast sensitivity deficits in patients with glaucoma.108–114 Additionally, measurement of the contrast sensitivity function has also been reported to be effective in revealing subtle visual losses in optic neuritis and multiple sclerosis,17,115–118 other optic neuropathies and cerebral abnormalities,115,119–123 amblyopia,124 and other conditions such as Alzheimer's disease,98 Parkinson's disease,115 and cystic fibrosis.125 The contrast sensitivity function can also be measured in infants and young children, with use of either electrophysiologic or behavioral techniques.126–131 There is accumulating evidence that these procedures are able to provide valuable clinical information about the functional visual status of infants and children. As these techniques become more refined and easier to use, they are increasingly being incorporated into clinical settings. In addition to providing useful information about the functional integrity of the visual pathways, there is also evidence that the contrast sensitivity function may be helpful in predicting the performance of various “real world” tasks, such as those involving the identification of objects such as aircraft132 or tanks,133 recognition of highway signs,134 reading ability,135–137 face recognition,138,139 and mobility skills of patients with subnormal vision.140 Thus, not only is the contrast sensitivity function useful for revealing subtle visual deficits associated with ocular disorders, but it also appears to be helpful in identifying problems that a patient is likely to encounter during daily activities. Recently, there have been several new approaches to the evaluation of contrast sensitivity using wall charts in a manner similar to that in which clinicians measure visual acuity. One of these methods is represented by the Vistech chart, shown in Figure 7.141 This chart has a series of five rows (A through E), each with a different spatial frequency. Each row consists of a group of nine circular targets containing a sinusoidal grating that is either vertical, tilted to the left, or tilted to the right. From the left side of the chart to the right, there is a successive reduction in the contrast of the grating. Patients are positioned at the 10-ft viewing distance and are asked to “read” each row from left to right and indicate the orientation of the grating. Another method of testing contrast sensitivity uses a typical letter chart for testing visual acuity, but with a lower contrast for the letters and the background. An example of a low-contrast version of the Bailey-Lovie chart is presented in Figure 8. Regan and Neima142,143 have also developed a series of visual acuity charts at several different contrast levels. Although normal observers will demonstrate a modest reduction in visual acuity for low-contrast targets, patients with early or subtle abnormalities will sometimes demonstrate quite profound reductions in visual acuity for low-contrast targets, in spite of good visual acuity for standard high-contrast letters. Thus, there is reasonable evidence that this approach to contrast sensitivity testing is able to reveal early disturbances of the visual pathways that are not revealed by standard visual acuity testing. The advantage of a low-contrast acuity chart in a clinical setting is that it is a test procedures that is highly familiar to most patients.
A third method of testing contrast sensitivity using a chart has been introduced by Pelli, Robson, and Wilkins and is known as the Pelli-Robson chart.144 As shown in Figure 9, the chart consists of letters of a fixed size that vary in contrast. Each line consists of six letters, with the three leftmost and three rightmost letters having the same amount of contrast. The patient reads the chart in a manner similar to that for a standard visual acuity chart, and the minimum contrast at which the letters can be detected is recorded. Several studies have reported that this method of testing contrast sensitivity is highly reproducible145–147 and is capable of detecting functional losses that are not evident with standard visual acuity testing. The Optic Neuritis Treatment Trial used the Pelli-Robson charts for evaluation of contrast sensitivity as an outcome measure for comparison of treatment groups.17 The use of sinusoidal gratings versus low-contrast letters for the measurement of contrast sensitivity is controversial. A number of arguments supporting each approach have been presented.148–150 Irrespective of the advantages and disadvantages of each method, it is important to note that different results are sometimes obtained by the two procedures. Different stimulus conditions in observers with normal visual function can result in different contrast sensitivity measures being obtained by gratings versus low-contrast letters.37 In addition, patients with certain visual disorders can produce different contrast sensitivity values for gratings versus letters.36 |
VISUAL FIELD SENSITIVITY | |||||
Perimetry and visual field testing are also tests of visual function that
are commonly performed in a clinical setting. The basic technique is
founded on the increment or differential light threshold, which refers
to the minimum amount of light that must be added to a stimulus for
it to be distinguished from a uniform background field. The relationship
between the increment threshold (ΔL) and background luminance (L) is
shown in Figure 10 on a log-log plot. At low levels, there is no influence of the background
luminance on the increment threshold (i.e., a minimum amount of light is needed to detect the stimulus). In this region, the
increment threshold is constant (ΔL = C [a constant]), as
indicated by the horizontal portion of the curve in Figure 10. At higher levels, the increment threshold increases as a function of
higher background luminances. In this region, a doubling of the background
luminance requires a doubling of the stimulus increment for it to
be detectable. This is denoted in Figure 10 by the portion of the curve with a slope of 1, which indicates that ΔL/L = C. This
relationship (ΔL/L = C) is known as Weber's
law or the Weber fraction in honor of the German scientist who was
the first to describe it in 1832.151 The background luminance that is typically used for perimetry and visual field testing is 31.5 apostilbs (10 candelas per square meter), which is approximately one third of the way up the portion of the curve in Figure 10 that has a slope of 1 (Weber's law). This is important, because changes in pupil size or other factors that affect retinal illumination will equally influence the background and stimulus, and thus will shift sensitivity up and down the Weber's law portion of the curve. Because ΔL/L = C in this region, increment threshold sensitivity remains unaffected. However, in some situations (e.g., elderly glaucoma patients with lenticular opacities and iatrogenic miosis), retinal illumination may be reduced to the nonlinear portion of the curve, confounding interpretation of responses. The increment (differential) threshold varies as a function of visual field location. With the standard background luminance of 31.5 apostilbs (10 cd/m2), the fovea has the lowest increment threshold. There is a progressive increase in the increment threshold with increasing eccentricity from the fovea. By convention, representation of visual field data is presented in terms of sensitivity, which is the reciprocal of threshold. In this view, the highest sensitivity (for the standard background luminance) is at the fovea, with a progressive decrease in sensitivity with increasing peripheral eccentricity. A three-dimensional representation of visual field sensitivity for a normal visual field is shown in Figure 11. Note that visual field sensitivity and the total extent of the visual field are greater for the temporal visual field than for the nasal visual field, and sensitivity and visual field extent are greater for the inferior visual field than for the superior visual field. Because of its characteristic shape, this representation of visual field sensitivity has previously been referred to as the “hill of vision” or the “island of vision in the sea of blindness.” It is important to keep in mind that visual field locations are inversely related to retinal locations. Stimuli in the superior visual field project to the inferior retina and vice versa. Similarly, stimuli in the temporal visual field project to the nasal retina and vice versa. Both the background luminance level and the size of the target used to perform visual field testing can affect the shape of the “island of vision.”152,153 In general, the slope of the visual field sensitivity contour is steeper for higher background luminances and/or smaller targets, and it is flatter for lower background luminances and/or larger targets. At very low background luminances, the fovea becomes less sensitive than surrounding regions, producing a contour resembling a “volcano of vision in the sea of blindness.” Three main techniques are used clinically to perform visual field testing: kinetic perimetry, static perimetry, and suprathreshold static perimetry. The most common form of kinetic perimetry uses a Goldmann perimeter, which consists of a white hemispheric perimeter bowl of uniform luminance onto which a small illuminated stimulus is projected. The patient's eye to be tested is aligned to the center of the hemispheric bowl. With the nontested eye occluded with an eye patch, the patient is instructed to maintain steady fixation on a small central target. The examiner is able to monitor the observer's eye position and fixation behavior by means of a telescope viewer. A stimulus of a selected size and luminance is projected onto the perimeter bowl and is moved from a nonseeing region in the far periphery toward the fixation point at a rate of approximately 4° per second.154 The patient presses a response button when the stimulus is first detected, and this location is marked on a chart. This procedure is repeated for a series of meridians around the visual field, each kinetic scan directed from a nonseeing region toward the fixation point. These locations of equal sensitivity are connected by a curved line referred to as an isopter. Because of the characteristic “hill of vision” shape of the visual field sensitivity profile, it is possible to evaluate different regions of the visual field by changing the size and luminance of the moving stimulus. To examine the far periphery, one would use a large, high-luminance moving stimulus that is readily detectable by locations with low sensitivity. Reducing the size or luminance of the stimulus makes it less detectable; therefore, the stimulus must be moved nearer to the central fixation point before it is detected by more sensitive regions of the visual field. Typically, a number of isopters are determined by the use of different stimulus sizes and luminances. Localized areas of nonseeing inside of isopters are called scotomas and are represented by areas that are solidly filled in on the visual field chart. The technique for performing quantitative kinetic perimetry is a highly acquired skill, requiring instruction and practice. A detailed discussion of kinetic perimetric technique is beyond the scope of this chapter. However, Anderson's book provides an excellent and thorough discussion of the intricacies of kinetic perimetric testing.155 A thorough kinetic perimetry evaluation of the visual field typically requires about 15 minutes for normal eyes, and up to 20 to 30 minutes for eyes with complex visual field loss. Representation of the results of kinetic perimetry is typically in the form shown in Figure 12A, which depicts the findings for a normal left eye. Figure 12B depicts the three-dimensional representation of the hill of vision and shows its relationship to the isopter plot for kinetic perimetry. Basically, the visual field representation for kinetic perimetry is a type of topographic relief map of the hill of vision. Pathologic changes to the visual pathways are denoted by specific patterns of visual field sensitivity loss. On the kinetic perimetry plot, sensitivity losses are depicted by displacement or constriction of isopters from their normal shape or the presence of localized scotomas within isopters.
Static perimetry, as its name implies, uses a stationary target whose visibility is altered by adjusting its luminance. In recent years, static perimetry has been popularized through its use by automated perimeters,156 and automated static perimetry is now considered to be the “gold standard” for visual field testing. With the fellow eye occluded, the observer is instructed to maintain steady fixation on a small central stimulus, and to respond by pressing a button each time a flash of light is detected in the peripheral visual field. If a stimulus is seen, its luminance is reduced for the next presentation; stimulus luminance is increased for the next presentation if it is not seen. A staircase or “bracketing” procedure is used to determine the increment or differential threshold, with a reduction in the size of the luminance increase or decrease each time there is a transition from seeing to nonseeing, or vice versa (i.e., a threshold “crossing”). For automated static perimetry, two threshold crossings are typically employed. Single stimulus presentations are pseudo-randomly distributed throughout the visual field so that the observer does not have any cues as to the location of the next stimulus, thus reducing the likelihood of eye movements. The status of staircase procedures for a number of locations (typically between 50 and 80) is thus tracked by the internal computer system. An equally-spaced grid of test locations is the most popular stimulus presentation pattern employed by automated perimeters. There are several advantages to the use of automated static perimetry.157 First, the procedures and test conditions are standardized, making it possible to compare results from one visit to another or from one clinic to another. Second, on the most commonly used automated perimeters, the test results of an individual are automatically compared to an internally stored age-related normative database. This permits a point-by-point determination of whether visual field sensitivity is within normal limits. Third, summary statistics, known as visual field indices, are generated to give an overall evaluation of diffuse or widespread sensitivity loss, localized sensitivity loss, losses that adhere to specific patterns associated with glaucoma and other disorders, and the degree of response fluctuation exhibited by the observer. Finally, a number of “catch” trials are included in the test procedure to check for fixation losses, false-positive responses, and false-negative responses. These are known as reliability indices. Representation of the results of static perimetry is usually in the form of a grey-scale graphical form, as depicted in Figure 13A. Areas of high sensitivity are denoted by lighter shadings, and areas of low sensitivity are represented by darker shadings. The relationship of the grey-scale representation to the three-dimensional “hill of vision” is shown in Figure 13B. As with the isopter visual field representation for kinetic perimetry, the grey-scale representation for static perimetry is designed to reflect the topographic characteristics of the hill of vision to reveal areas of normal and reduced sensitivity to light.
Suprathreshold static perimetry is a rapid screening procedure for evaluation of the peripheral field of view.156 Rather than determining a threshold, suprathreshold static perimetry presents targets that are above threshold at a number of visual field locations. It therefore has very limited quantitative capabilities, although it is able to give a quick determination of whether there are any abnormalities in the visual field and, if so, their location. Three basic procedures have been used for suprathreshold static perimetry. The first merely presents a stimulus of a fixed luminance at each of the visual field locations. This is a simple procedure that can be performed quickly, but because visual field sensitivity varies with eccentricity, some visual field locations will be more suprathreshold than others. A second method adjusts the suprathreshold stimulus to a fixed amount above the average age-adjusted normal threshold. Thus, the stimulus luminance is different for various eccentricities and is presumably a fixed amount above the normal threshold. This is a better screening procedure, but it is a bit more complex. A third method performs a threshold evaluation for one to four visual field locations; it uses this information to estimate threshold values for the entire visual field and adjusts the suprathreshold stimulus to a fixed amount above this value for each visual field location. This is a more accurate screening method, but it requires more time and is more complicated. It has been demonstrated that automated suprathreshold static perimetry can be used to screen very large populations and provide useful information about the status of the visual field.158–161 There are several aspects of perimetry and visual field testing that are of clinical value. First, it is a means of detecting early functional deficits produced by a wide variety of eye diseases, many of which have their initial effects on peripheral vision while having little or no effect on foveal vision. Glaucoma is perhaps the most notable of these types of disorders. Second, the patterns of sensitivity loss in the visual field are of value for differential diagnostic purposes. Specific features of localized regions of reduced sensitivity can help define the locus of pathology in the visual pathways and the type of disease entity that is likely to be present. Third, patients are often unaware of peripheral vision loss, especially if it has progressed gradually. In a study of 10,000 driver's license applicants, it was found that approximately 3% had some loss of peripheral vision in one or both eyes, and more than half of them were unaware that they had any problem with their vision.159 Glaucoma, retinal disorders, and cataracts were the most common reasons for their peripheral vision loss. And finally, the peripheral visual field is important for driving159,162 and other mobility tasks.140 It is beyond the scope of this chapter to provide a detailed description of all the patterns of visual field loss that are associated with various disorders of the visual system. There are several excellent sources available that provide numerous examples of visual field deficits associated with lesions at specific loci in the visual pathways.155,163–167 Harrington's book is a particularly good source for examples of visual field loss and clinicopathologic correlations,164 and Frisen's book includes a comprehensive bibliography of relevant perimetry literature.163 In general, the patterns of visual field sensitivity loss correspond to the anatomic arrangement of nerve fibers at various sites in the visual pathways. Several examples, based on automated static perimetry with use of a grey-scale representation, are shown in the following figures. An arcuate nerve fiber bundle defect that is typical of glaucoma is presented in Figure 14. Note that the area of sensitivity loss “respects” the horizontal meridian in the nasal visual field (i.e., the defect stops abruptly at the horizontal midline). In addition, the defect “fans” out in an expanding arc from the blind spot. The blind spot corresponds to the location of the optic nerve head and is centered at 15° eccentricity on the horizontal meridian in the temporal visual field. The pattern of sensitivity loss shown in Figure 14 is consistent with damage to ganglion cell nerve fibers at the optic nerve head, as produced by glaucoma.
Figure 15 presents an example of a bitemporal visual field deficit that is characteristic of lesions of the optic chiasm. At the chiasm, fibers from the nasal retina of each eye cross over to join fibers from the temporal retina of the other eye. Damage to these crossing nasal retinal fibers produces visual field loss in the temporal visual field of both eyes. Because the nasal and temporal retinal nerve fibers separate along the vertical midline, visual field loss resulting from chiasmal lesions demonstrates a “vertical step,” which is a distinct transition in sensitivity that occurs along the vertical meridian. The example shown in Figure 15 is the result of a pituitary tumor compressing the optic chiasm. The central 30° visual field representation for the left eye shows a nearly complete loss of sensitivity for the temporal visual field. In the right eye, there is nearly complete loss of sensitivity in the superior temporal quadrant, with partial sensitivity loss in the inferior temporal quadrant. Note that in both eyes, the areas of sensitivity loss “respect” the vertical meridian, demonstrating an abrupt transition in sensitivity at the vertical midline.
An example of a right homonymous hemianopsia, produced by a lesion to the optic radiations beyond the chiasm, is shown in Figure 16. The term homonymous refers to the fact that the defect is on the same side of vision, the right side, in both eyes. Thus, it is the temporal visual field (from nasal retinal fibers) in one eye and the nasal visual field (from temporal retinal fibers) in the other eye. The term hemianopsia refers to the fact that the defect involves half of the visual field, or a hemifield. A deficit involving only a quarter of the visual field (quadrant) is referred to as quadrantanopsia. Note that again the vertical meridian is respected in both eyes. This vertical step is a characteristic feature of lesions affecting the chiasmal and postchiasmal visual pathways.
There are many additional patterns and features of visual field loss that are associated with disorders at specific sites in the visual pathways and particular disease entities. The examples presented above are just a few illustrative examples of these patterns. It is beyond the scope of this chapter to provide a detailed account of patterns of visual field sensitivity loss associated with all types of visual disorders. Several excellent sources are available for those readers interested in a more extensive description of specific types of visual field sensitivity loss patterns.155,163–167 In recent years, a number of new test procedures have been adapted for clinical visual field testing to examine properties other than achromatic increment threshold characteristics of the peripheral field of view. The rationale underlying the development of these test procedures is that they may reveal information about the pathophysiology of specific ocular disorders, as well as provide earlier detection of functional changes associated with particular eye diseases. A brief review of many of these new perimetric test procedures has recently been published.168 Short wavelength automated perimetry (SWAP) uses the two-color increment threshold technique of Stiles169 to isolate and measure the activity of short-wavelength (blue) sensitive mechanisms. This is achieved by using a high-luminance yellow background and a large (Goldmann size V) blue stimulus. Recent longitudinal investigations have shown that SWAP reveals earlier and more extensive losses in glaucoma and is able to predict the development and progression of glaucomatous optic nerve damage.170–181 High-pass resolution perimetry (HRP) uses a series of “ring” targets of different sizes that are displayed at various locations on a video monitor. The ring targets are thus similar to visual acuity optotypes such as the Landolt C and consist of a dark ring surrounding a light center region. The HRP ring targets are generated by performing a high-pass spatial frequency filtering of the stimulus. Frisen designed the stimuli in this manner so that the thresholds for detection and resolution of the stimulus are essentially identical (i.e., the ring can be resolved at the same target size at which the stimulus is first detected).163,182 Recent studies of patients with glaucoma, retinal disease, and various neuro-ophthalmologic disorders of the optic nerve and chiasm indicate that HRP provides useful information about early functional losses.163,182–187 Pattern discrimination perimetry (PDP) is a new visual field test procedure developed by Drum that also uses a video monitor.188 The test consists of the detection of an alternating checkerboard pattern of black and white dots on a dynamic background of random black and white dots. The visibility of the stimulus is varied by changing the percentage of dots adhering to the alternating checkerboard pattern (coherence). Thus, the coherence of the stimulus can be varied from 100% (complete checkerboard) to 0% (completely random). Studies of PDP to date indicate that it detects a greater number of deficits than does standard automated perimetry in glaucoma patients.188–193 Flicker and temporal modulation perimetry (TMP) are procedures that evaluate the temporal processing characteristics of the visual system at various locations throughout the visual field. Lachenmayr has introduced one form of flicker perimetry that measures critical flicker fusion (CFF) for a 100% modulation target.194–196 This method thus measures the highest temporal frequency for which flicker can be detected. Another procedure described by Casson measures the amount of flicker modulation or contrast needed to detect flicker for low (2 Hz), medium (8 Hz), and high (16 Hz) frequencies of flicker.197,198 Recent studies indicate that both techniques are useful for early detection of glaucomatous damage. TMP has additionally been shown to be predictive of future glaucomatous visual field loss, and has also been effective in demonstrating early functional losses in other optic neuropathies.199 |
CONCLUSION |
New tests of visual function are constantly being developed and introduced into the clinical ophthalmologic environment. To be useful as a clinical evaluation method, a new test must evaluate fundamental sensory properties of vision and have “face validity,” demonstrate high sensitivity and specificity for distinguishing subtle abnormalities from normal responses, demonstrate high reliability and reproducibility, be robust to variations in test administration and test conditions in a clinical setting, be easy to administer and interpret, and offer advantages over existing test procedures. Very few evaluations of visual properties are able to satisfy all of the above criteria for a good clinical test of visual function. Visual acuity, contrast sensitivity, and evaluation of the peripheral field of view are three techniques that have established a role in the clinical evaluation of visual function. |
ACKNOWLEDGMENT |
The author and editors wish to acknowledge Dr. Lawrence F. Jindra, the author of the previous Chapter 17, “Evaluation of Vision.” Some of his material has been incorporated into the present chapter. |