Abstract
Purpose::
In microperimetry, the locations of visual stimuli are reported on a retinal image. The term microperimetry came to represent retinal perimetry even though perimetry with as small or smaller targets can be performed in standard perimetry. The reason that this terminology was established is the assumption that retinal perimetry provides more accurate placement of targets, in relation to fundus features. This was indeed the case in early scanning laser ophthalmoscopes (SLO) where the imaging laser was also used to create the perimetric stimuli and therefore target position was known exactly. Most microperimeters today use different systems for stimulus presentation and retinal imaging. Reported locations on the retinal image will be incorrect when not spatially aligned. We report generic psychophysical methods for confirming spatial alignment of microperimeters.
Methods::
Two techniques were developed and evaluated using normally-sighted observers and Nidek MP-1 instruments. (1) The physiological blind spot was identified using small stimuli. The center of the optic disk identified by locating the apparent edge of the disk in the retinal image was compared to the center of the blind spot. (2) Angioscotomata of at least two major retinal vessels with orthogonal orientations were mapped. The locations of those angioscotomata were compared to the locations of the vessels in the retinal image. In addition to these techniques, we compared the locations of the fovea of 16 normally-sighted observers to published SLO-derived normal population data based on optic disk location. The effect of head tilt on computed foveal location was examined.
Results::
Repeatable spatial alignment errors of 0.5 degree or more could be found using both techniques. Measurement errors associated with different operators, subjects and images were less than about 0.2 degrees. Differences between our small population sample and previous studies were explained by spatial misalignment. Even small tilts of the head produced apparent changes in foveal location.
Conclusions::
These techniques for assessment of spatial alignment could be applied to any microperimeter. Only microperimeters that image the retina and present stimuli using the same system, as found in some SLOs, are not at risk of spatial misalignment. Mislocalization of the foveal location due to spatial misalignment and head tilt may affect studies of macular lesions such as AMD.
Keywords: perimetry • clinical research methodology • imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound)