Abstract
Abstract: :
Purpose: The Rodenstock Scanning Laser Ophthalmoscope (SLO) images the fundus and simultaneously presents stimuli directly onto the retina. The purpose of this development was to establish automatic online positioning of stimuli in the SLO independent of eye movement inaccuracies in patients with unsteady fixation. Methods: We developed new microperimetry software for the SLO that tracks features of the retina in real time and automatically places stimuli in preselected positions. A still SLO fundus image is used to choose location, intensity (dB) and size (equivalent to Goldmann I–V) of the test stimulus. Presentation time was 250 ms. A digital frame grabber board is used to analyze each video frame in real time. A tracking algorithm finds a user–defined feature of interest (e.g. a vessel branching) 25 times per second to detect retinal movements. If an eye movement has occurred, the placement of the test stimulus is shifted automatically to compensate for the eye movement. We investigated 10 patients with central scotomas due to age–related macular degeneration and compared our method with the standard Rodenstock Scotometry software requiring manual compensation for eye movements. Results: We provide video evidence that the automatic tracking function works reliably and decreases investigation time by at least 25% using the same number of test stimuli. The positioning of the stimuli is simplified and more accurately because of the usage of a still fundus image. The extent of the scotoma was found to be identical in both methods. Despite of eye movements, angio scotomas of retinal vessels near the optic nerve head can be detected. Conclusions: Our new software allows monitoring of disease and its progression and is of excellent value in patients with macular lesions that induce a central scotoma and unstable fixation.
Keywords: age–related macular degeneration • eye movements: recording techniques • visual fields