Abstract
Purpose :
Live retinal imaging aids assessment of visual function. A lower cost device can show retinal location of visual stimuli during micro-perimetry, scotoma mapping, or fixation stability tasks.
Methods :
To demonstrate the potential range of visibility of stimuli on the fundus, 11 subjects 45.6 + yr were tested without mydriasis with a broad range of age and fundus pigmentation: 2 African/African American, and 9 Caucasians or bi-racial. We used a digital light ophthalmoscope (DLO) to illuminate 26 deg of retina at < 24 microwatts. The red LED raster of a DLP (Lightcrafter 4500, TI) was presented in sequential stripe pattern at 15 or 20 Hz. A CMOS camera with a rolling shutter synched to the illumination stripes improved contrast by reading only when specific lines were illuminated. Black stimuli drawn on the retina by the DLP were viewed live by the subject. The operator viewed the retina to position and observe the subject’s fixation, using also a second live image that showed the position of a graphics overlay of the stimulus on the retina. The size of the fixation pattern and the selection of task and visual stimuli were under computer control. Image alignment after the completion of a trial corrected stimulus location, similar to previous microperimetry techniques.
Results :
Over a wide range of fundus pigmentation, major retinal landmarks were visible in all subjects. Larger black on red stimuli, e.g. Goldman III targets, were readily visible in live images, but not when the black stimulus overlaid a dark fundus feature. Decreased visibility was due to contrast, not numbers of pixels (1662). For a pale to medium fundus over a bright feature, the Michelson contrast = .12 Stimulus location was determined from image alignment, since the software kept track of the stimulus position. Variation in retinal location is available, e.g. for 3 sequential images captured during a 200 ms target presentation.
Conclusions :
Visual function testing with visualization of the stimulus position with respect to the fundus is readily performed over a wide range of fundus pigmentations. Behavior calibration of each stimulus is unnecessary as long as there is at least one image with sufficient visibility of the fixation target or other landmark to use to re-align the images. This allows for investigation of strategies to use scanning to enhance visibility of targets for patients with low vision.
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.