Perimetry provides more detailed evaluation of the visual field. In both static and kinetic techniques, the visual field is analyzed for areas of decreased sensitivity, both in location and in degree. In static testing, stimuli turn on and off at designated points within the region of the visual field being tested. In kinetic testing, a stimulus moves from a nonseeing to a seeing area of the visual field to determine the location at which it is consistently detected by the patient. All points of equal sensitivity for a specific stimulus are connected to form an isopter, which represents the outer limit of visibility for that stimulus. Analysis of several isopters (plotted with different stimuli) produces a “contour map” of the island of vision.
Kinetic perimetry (eg, Goldmann perimetry, the kinetic program on an Octopus perimeter [Haag-Streit, Köniz, Switzerland]) can be used to evaluate the entire visual field. Stimuli of varying sizes and intensities are moved along each radial meridian from a peripheral to central location. Typically, 2 or 3 isopters are plotted. Varying the stimulus size, intensity, and location can delineate the depths and borders of defects. Kinetic perimetry requires a skilled and knowledgeable perimetrist who can interact with patients to elicit optimal cooperation.
Automated static perimetry
Automated static perimetry is considered the gold standard for evaluating visual field defects. Although this method is particularly difficult to use with older or inattentive patients, it possesses numerous advantages over manual kinetic perimetry techniques:
standardized testing conditions, which improve serial and inter-institutional comparisons of results
less dependence on technicians
numerical data amenable to statistical analysis for comparisons and clinical studies
results amenable to electronic data storage
Most automated perimeters use static stimuli that are similar in size to the kinetic perimeter size III stimulus. The perimeter randomly presents stimuli at predetermined locations within a specified region of the visual field. Because nearly 80% of the visual cortex correlates to the central visual field, testing the central 24° or 30° of the visual field is typically adequate for detecting most visual defects (Fig 3-5). It is also critical to measure the foveal threshold (normal range 32–40 dB) because it estimates central visual function. Because of the 6° test spacing of the 30-2 and 24-2 programs, 10-2 perimetry may better delineate a small central or paracentral scotoma.
Figure 3-5 Kinetic versus automated perimetry. Diagrammatic representation of the extent of the visual field evaluated by kinetic perimetry versus the 30° central program in automated static perimetry. The largest isopter in kinetic perimetry testing extends 90° temporally and 60° in other quadrants; typical automated static perimetry evaluates only the central 30°.
(Courtesy of Anthony C. Arnold, MD.)
The stimuli vary in brightness, and patient responses determine the minimum visible stimulus at each location—the sensitivity threshold. This threshold is defined as the intensity of the dimmest target identified 50% of the time at a given location. For each region tested, the report displays the threshold value in decibels (on a logarithmic scale of intensity, measuring attenuation from the maximum stimulus of the perimeter). For a given stimulus location, a higher value indicates that the patient can see a dimmer stimulus, reflecting greater visual sensitivity at that location. The measured values are not absolute numbers and are not equivalent among perimeters because the machines have different maximum intensities, backgrounds, and durations of presentation.
A symbolic representation of the threshold values, the grayscale map, depicts an overall topographic impression of the visual field data by using darker symbols for low-sensitivity points and lighter symbols for high-sensitivity points. The computer program interpolates between tested points to provide a user-friendly picture (Fig 3-6). For clinical interpretation, the perimeter calculates for each value the statistical probability that the value falls outside the normal range among age-matched control subjects; the results are placed in a total-deviation plot. Optic neuropathy may cause substantial total-deviation depression with few or no pattern-deviation abnormalities. Because ocular media abnormalities (eg, refractive error, cataract) may depress the sensitivity of the entire visual field, the program produces a pattern-deviation plot by determining the sensitivity values for all points shifted (by the seventh-highest point) and reanalyzes them based on age-expected values. This reanalysis compensates for the overall sensitivity depression, allowing recognition of abnormal patterns (eg, scotomata, arcuate defects, homonymous defects) that might have been otherwise masked. Abnormal values are depicted topographically according to statistical probability: Darker squares represent higher probability, and lighter squares represent lower probability.
The long duration and repetitiveness of the original full-threshold perimetry test can fatigue patients, reducing the accuracy of the test results. Use of the Swedish interactive threshold algorithm (SITA) shortens the time needed to complete the full-threshold test by half but maintains the accuracy necessary for validity. (See BCSC Section 10, Glaucoma.)
The reliability of perimetry test results is assessed by identifying the following patient response characteristics:
false-positive response rate: how frequently the patient signals when no target is displayed (The acceptable rate is typically <25% on threshold testing and <15% on SITA testing.)
false-negative response rate: how often the patient fails to signal when a target brighter than the previously determined threshold for that spot is displayed (The acceptable rate is typically <25%, but the rate increases in regions of true visual field loss, because the patient is unable to reproduce responses accurately.)
fixation losses: how often the patient identifies the stimulus in the previously plotted physiologic blind spot (an unexpected response), indicating that the eye is not aligned with the fixation target
Global indices are calculated to help determine changes in sensitivity over time. Such indices include a center-weighted mean of the sensitivity depressions across all points (ie, mean deviation) and different methods of addressing localized defects (eg, pattern standard deviation, corrected pattern deviation, loss variance).
Figure 3-6 Report from a Humphrey 30-2 automated static perimetry program, with explanations of statistical analysis, grayscale, and probability plots (red type).
(Courtesy of Anthony C. Arnold, MD.)
Barton JJS, Benatar M. Field of Vision: A Manual and Atlas of Perimetry. Totowa (NJ): Humana Press; 2003.
Bettis DI, Johnson CA. Updated on automated perimetry. Focal Points: Clinical Practice Perspectives. San Francisco: American Academy of Ophthalmology; 2016, module 12.
Excerpted from BCSC 2020-2021 series: Section 5 - Neuro-Ophthalmology. For more information and to purchase the entire series, please visit https://www.aao.org/bcsc.