Shortwavelength Automated Perimetry

Short-wavelength automated perimetry (SWAP) tests the detection of a large blue stimulus against a yellow background. The targets are retinal ganglion cells with blue-yellow color opponency, which account for 1% of the total population. These blue-yellow ganglion cells form a specific class that project to the intralaminar-koniocellular layers of the lateral geniculate nucleus (15). They have large dendritic fields and axons, nearly as large as those of the magnocellular cells, and may thus number among the large fibers preferentially lost in pathologic studies of glaucomatous damage to the optic nerve.

A number of studies have shown that SWAP detects glaucomatous defects earlier than conventional perimetry (14). How ever, it may be more variable between tests, and its thresholds may be adversely affected by posterior subcapsular cataracts.

The latest models of automated perimeters have the capability of performing SWAP perimetry, so this technique is already a clinical reality.


The stimulus in high-pass resolution perimetry is a bright ring (25 candles [cd]/m2) sandwiched by two dark rings (15 cd/m2), a pattern that can be obtained by high-pass spatial filtering of a single luminant ring (Fig. 4). The luminance averaged over all rings is the same as the background; hence, contrast rather than brightness is being used to define the target. Instead of varying contrast, the test varies the size of the ring, to find the smallest detectable ring in a given region. The special feature of these stimuli is that the threshold for detection is close to the threshold for resolution, which is the ability to perceive the two dark lines as separate elements. Thus, the width of the bright ring is the key feature in detection (16). There are good theoretical grounds and empiric evidence that perceiving this fine spatial detail depends on the parvocellular ganglion cell population (17,18).

High-pass resolution perimetry appears to be as good as conventional automated perimetry at detecting glaucomatous defects, but not any better (19-21). A similar conclusion is probably valid for neuroophthalmologic conditions (22). The clinical appeal of this new perimetry lies in some practical advantages. It takes 50% less time, patients seem to feel more comfortable with it, and it has less intra-test variability (22,23). Age, stimulus location, and pupil diameter also do not adversely affect variability (24), but the results may be more vulnerable to poor focus from inadequate refraction or cataract (16). A commercial system is available but requires its own hardware.


Two main strategies of motion perimetry have been used. One is to determine for a single spot or line the smallest position shift that is detectable as stimulus motion (minimum displacement threshold) (25). The second is to present a swarm of moving dots. Some, belonging to a noise pool, move randomly; others, in a signal pool, move in a common direction. The threshold is the lowest ratio of signal-to-noise dots at which the subject can accurately guess the signal direction (26). Motion is thought to be processed more by magnocellular than parvocellular cells, though this selectivity does not likely mean exclusivity.

Studies of patients with glaucoma or ocular hypertension suggest that motion perimetry might be more sensitive than conventional perimetry for nerve fiber bundle defects in these conditions (25-28). However, stimulus size and duration may be critical variables in determining the sensitivity of the technique (28). There is a similar suggestion of better sensitivity to arcuate defects in IIH (29), which may share with glaucoma a pathologic effect of increased pressure at the optic nerve head. Although there are some claims of immunity to the effects of refractive blur and cataracts, defocus does affect foveal motion thresholds in complex ways that depend on stimulus displacement and velocity (30).


This unique type of perimetry actually tests for an illusion. When a low spatial frequency (<1 cycle/°) sinusoidal grating is rapidly flickered (>15 Hz) in counterphase (meaning that the white peaks become the dark troughs and vice versa), there is an illusion

Was this article helpful?

0 0

Post a comment