THE NORMAL VISUAL FIELD
The normal visual field has been described as an island of vision in a sea of darkness. This island has a sharp central peak, corresponding to the fovea, with sloping sides. The sides are slightly steeper superiorly and nasally. The island of vision extends roughly 60° superiorly and nasally, 75° inferiorly, and 100° temporally ( Fig. 8-1 ). The actual topography (sensitivity of various parts) of the island depends on the level of light adaptation of the retina. The peak is most sensitive when the retina is light adapted. The edges of the island have poor light sensitivity, so stimuli must be up to 3500 times (3.5 log) more intense to be perceived. If the retina is fully dark adapted, the cones of the fovea (center of the island) are less sensitive than the rods of the periphery.
Visual field testing is usually done in the photopic (light-adapted) or mesopic (partially light-adapted) state. Thus in the normal visual field examination, the fovea is the most sensitive point tested and represents the peak.
VISUAL ACUITY VERSUS VISUAL FIELD
Visual acuity measurement tests the resolving power of the retina for objects of distinct form. Static visual field measurement tests a more primitive retinal function – differential light sensitivity. Differential light sensitivity is the measure of the ability of the retina to distinguish a stimulus that is some degree brighter than the background illumination.
TERMINOLOGY AND DEFINITIONS
Fixation. That part of the visual field corresponding to the fovea centralis. Also, the ability of patients to keep their eyes directed at the center of the visual field apparatus. Patients with poor fixation move their eyes repeatedly and produce an unreliable visual field test result.
Central field. That portion of the visual field within 30° of fixation.
Bjerrum’s area (arcuate area). That portion of the central field extending from the blind spot and arcing above or below fixation in a broadening path to end at the horizontal raphe nasal to fixation. Bjerrum’s area usually is considered to be within the central 25° of the visual field. This part of the visual field is quite susceptible to glaucomatous damage (see Fig. 10-9 ). Bjerrum’s area does not include non-specific peripheral depression that is commonly seen along the uppermost border of automated visual field charts. These defects may appear to arc because of the placement of test points, but they do not constitute a classic arcuate scotoma ( Fig. 8-2 ; see also Fig. 10-2 ).
Fig. 8-2
The dark areas at the top of this printout are artifacts. The patient has a normal field.
Peripheral field. That portion of the visual field from 30° to the far periphery. The shape of the normal peripheral field is governed by the shape and structures of the face.
Kinetic perimetry. Visual field test wherein the intensity and size of the stimulus are held constant while the stimulus location is moved.
Static perimetry. Visual field test wherein the position of the stimulus is held constant while the stimulus intensity is varied.
Isopter. The outline of a contiguous area of the visual field capable of perceiving a given stimulus. The isopter is most often used to define an area outlined by a given stimulus in kinetic perimetry.
Threshold. At a given retinal point, the intensity of a stimulus that is perceived 50% of the times it is presented.
Fluctuation. The variability in visual field measurement when tests are repeated over time.
Short-term fluctuation. The variability within a field during the time of its measurement. The test–retest interval is short: typically a few seconds or minutes.
Long-term fluctuation. The variability between two visual fields performed sequentially on the same eye that cannot be attributed to pathologic change. The test–retest interval is typically days to months or longer.
Short wavelength, automated perimetry (SWAP). Visual field test in which short wavelength-sensitive (blue) cones are isolated by using blue light stimuli projected on a yellow background. Also called blue-on-yellow perimetry.
Frequency doubled perimetry. Visual field test in which stimuli are alternating high-frequency contrast bands.
Depression. A reduction in expected (normal) sensitivity.
Scotoma. A localized defect or depression within the visual field.
Absolute defect. A field defect that persists when the maximum stimulus of the testing apparatus is used. The normal blind spot is an absolute scotoma.
Relative defect. A field defect that is present to weaker stimuli but disappears when tested with brighter stimuli. A defect that is not absolute (see Fig. 10-4 ).
Candela per square meter (cd/m 2 ). The international unit of luminance.
Apostilb. 0.1 millilambert = 3.183cd/m 2 .
Log unit. Logarithm base 10 of the luminance in apostilbs.
Decibel. One-tenth of the log unit.
THEORY OF VISUAL FIELD TESTING
The purpose of visual field testing is to define the topography of the island of vision to recognize any variation from normal. It is used to detect abnormalities and to follow abnormalities while the patient is under observation or treatment. The visual field is tested by adapting the eye to the background luminance and then presenting a stimulus that is some degree brighter than the background at a given position in the field. The ability of the patient to perceive the stimulus may be tested kinetically, statically, or with some combination of the two techniques.
KINETIC PERIMETRY
Kinetic perimetry is typically performed manually by confrontation, on a tangent screen, or with a Goldmann perimeter. In kinetic perimetry, the stimulus usually is presented in the non-seeing periphery and moved at approximately 2° per second toward fixation until the patient first perceives it. The stimulus is subsequently moved to another meridian in the periphery out of view and advanced toward fixation again until the patient sees it. By repeating these maneuvers at approximately 15° intervals around 360° of the visual field, the examiner defines a series of points that can be connected to describe an isopter corresponding to the stimulus used (see Fig. 8-1 ). By decreasing or increasing the size or brightness of the stimulus, a smaller or larger isopter will be outlined. If the stimulus is presented into randomly selected areas of the visual field, the isopters will be slightly constricted and irregular compared with sequentially presented stimuli. Reproducibility may be greater with sequentially presented stimuli.
After initial detection, a scotoma can be defined more precisely with kinetic perimetry by placing the stimulus in the scotoma and moving the stimulus outward until it is perceived. This process is repeated in various directions until all edges of the scotoma have been defined. If the edges of the scotoma are sloping (the change from normal to abnormal regions within the field is gradual), a brighter stimulus will define a smaller scotoma, and a dimmer stimulus will define a larger scotoma. If the margins of the defect are steep, changing the stimulus size or intensity will affect the size of the scotoma only slightly.
STATIC PERIMETRY
In static perimetry, the test stimulus size usually remains constant throughout the test. For computerized full-threshold testing, each point in the visual field is evaluated by positioning the stimulus at a test point and varying the intensity until the threshold for that particular retinal location is defined ( Fig. 8-3 ). This process is repeated until all of the positions of the retina to be measured have been tested.
The more retinal positions tested, the more defects will be found and quantified. There is, however, a point of diminishing returns, at approximately 80 locations, wherein patient fatigue seriously reduces the accuracy and consistency of responses. Most computerized perimeters ( Figs 8-4 and 8-5 ) use static visual field testing techniques for their standard tests.
Alternatives and modifications to standard full-threshold testing of each retinal position have been devised to reduce the number of patient responses required without reducing the amount of information obtained at each testing session. Such alternatives include threshold-related testing and zone testing, as well as algorithms that use less precise bracketing to estimate the threshold. These methods generally produce results that are similar to, but can be somewhat more variable than, standard threshold determining strategies. The most widely used modification of standard repeat-bracketing threshold testing is the Swedish Interactive Testing Algorithm (SITA) program used on the Humphrey Field Analyzer, which adjusts the starting and ending points of the bracketing procedure during the examination based on the patient’s responses. This is done in a fashion that reduces redundancy, decreases testing time, and increases accuracy and patient acceptance without compromising the sensitivity and specificity of the test (see also p. 95).
THRESHOLD-RELATED TESTING
The ‘normal’ state of the visual field is a statistically determined figure obtained from the testing of many normal individuals of different ages, genders, etc. Each retinal location has a statistically determined ‘normal’ sensitivity range that can be expressed in decibels of stimulus intensity related to stimulus size, background intensity, and patient age. This sensitivity is not constant from patient to patient, or even within the same patient from test to test. Therefore, for a particular retinal location to have a strong possibility of being abnormal, its sensitivity should be reduced from normal by roughly two standard deviations of the mean of normal, or approximately 4 or 5 dB on average. This is conveniently expressed as twice the average short-term fluctuation (SF). It is similar to the traditional rule of thumb that suggests that significant measurement deflections are at least twice the baseline noise level. The average short-term fluctuation is lower at or near fixation than it is in the periphery, so a deviation of 4 or 5 dB centrally has a greater chance of representing a reproducible change in sensitivity than does a defect of similar depth in the periphery. A 4 dB depression in an area of the field that has a SF of 1.2 dB is more likely to represent pathology than a 10 dB depression in an area than has a SF of 8 dB, which can be the case at 24° or more from fixation.
In threshold-related testing ( Fig. 8-6 ), if the patient is presented with a stimulus that is roughly 4 dB brighter than the expected normal level for that retinal position and the patient sees it, the location is considered normal, and the stimulus is moved to the next position without measuring the threshold of the location precisely.
The disadvantage of this technique is that it only finds defects equal to or greater in depth than the suprathreshold stimulus used. This technique also provides no information regarding subtle variation in the contour of the field, which is important in recognizing early changes from normal. The rapidity of testing normal areas using the technique, however, allows a larger area of the retina to be examined. If defects are detected, they can be quantified with the full-thresholding strategy.
ZONE TESTING
Zone testing uses three levels, or zones, of stimulus intensity to locate and then quantify defects. The first zone is a suprathreshold stimulus 4 or 5 dB brighter than the anticipated normal threshold, as described in the section on threshold-related testing, above. If the patient sees this stimulus, the response is recorded as normal. If the patient fails to see the initial stimulus, a maximally bright stimulus is shown. If the patient sees this stimulus (but failed to see the initial, relatively dim, stimulus), the machine indicates a relative defect. If the patient fails to see either stimulus, the machine records an absolute defect. Responses can thus be grouped in three zones – normal, relative defect, or absolute defect. There are multiple variations on this theme that allow a greater number of zones to be defined, or for zones to be defined at different levels. The obvious disadvantage of this technique is that subsequent testing can only recognize major change because the difference between the test stimuli is great. The advantage is that it is fast.
SCREENING TESTS
Screening tests for visual field defects are available by manual perimetry and with most computerized perimeters. Unfortunately, they only detect rather large changes in the visual field. Most screening programs use a technique that recognizes defects that are greater than 4 or 5 dB below an expected level. As such, they may not detect early glaucomatous defects. Also, if a defect is found, the patient must undergo a full visual field examination to establish a baseline against which future change can be measured. Therefore, in persons suspected of having visual field defects, screening programs and other fast strategies that reduce examination time at the expense of evaluating the critical sensitivity area within 5 dB of threshold are of questionable value. Screening tests are used most appropriately to diagnose pathology rather than to follow or quantitate the degree of damage in a patient known or suspected of having disease.
OTHER STATIC TESTING TECHNIQUES
Single-stimulus level static testing was used commonly in some of the early computerized perimeters. Because the visual threshold slopes from the center to the periphery of vision, a single-stimulus intensity cannot be effective in testing a large area of the retina. This technique is useful only as a relatively crude screening method. A variety of innovative screening techniques has been developed to aid clinicians in obtaining the maximum amount of useful information in the minimum amount of time. A few of these bear mention for illustrative purposes.
Noisefield perimetry, also known as white-noise perimetry or campimetry, is performed by having the patient observe a computer screen ( Fig. 8-7 ) or home television set. The screen projects a ‘noise’ pattern of small (roughly 1–4 mm), irregularly shaped dark and bright spots oscillating at 50 Hz. Patients with localized defects notice the defective region as a smudge or blank area on the screen. In simple terms, they detect the noise pattern in normal regions of the field, and its absence in the abnormal areas is perceivable. Detection takes only seconds in alert patients, but not all patients can cooperate fully with the test requirements. The information gained is useful mainly for screening. Optokinetic perimetry is a novel approach to visual field screening in which the patient is presented a series of cards with static stimuli arranged in a set pattern. While maintaining steady fixation, the patient is asked how many spots she sees. Stimuli in defective areas of the field are not detected, and by evaluating the points missed during the test, the examiner can gain a fairly clear idea of the nature and extent of the field loss. This method is very fast, taking perhaps only 30–50% of the total time needed for screening fields conventionally. These methods are used rarely, if at all, in clinical practice.
