Abstract
Purpose :
To visualize retinal structures using a newly developed parallel line scanning ophthalmoscope (PLSO).
Methods :
A PLSO was built using a digital micromirror device (DMD) instead of traditional scanning mirrors to scan lines over the field of view (FOV). The DMD consists of 912 × 1140 micromirrors which can be individually switched on/off based on a programmed binary pattern. By switching on multiple (parallel) two-element wide lines in the DMD, the corresponding lines on the retina are imaged on a CMOS camera. After acquisition of each frame, the micromirrors are turned off and the mirrors for the next set of adjacent lines are turned on. This is repeated until the whole FOV is imaged. Confocal images are generated from the data by subtracting the maximum and minimum intensity values for each pixel in the sequence. The fovea and optic nerve head (ONH) of a healthy subject were imaged using 10° × 10° FOV at 100 Hz with 7 parallel lines resulting in a full image frame rate of 1.4 fps. The images were acquired through a dark-adapted pupil without any dilatation. The acquired data were processed, as mentioned earlier, into confocal images; but also non-confocal images were obtained by averaging all frames.
Results :
Figure 1A shows the imaged areas. In the non-confocal images (Fig. 1B&C), the corneal scattering is dominant and makes the retinal structures covered in haze. In the confocal images (Fig. 1D&E), confocality and contrast are improved. The foveal avascular zone and smaller blood vessels are visible in the fovea image (Fig. 1D). Also the quality of the ONH image is improved and many of the main features can be distinguished such as small blood vessels (Fig. 1E).
Conclusions :
The PLSO provided high contrast images of the fovea and ONH and detailed retinal structures could be observed. The DMD eliminates moving parts from the system and exposure time for each frame is potentially shorter than in full-field imaging, which reduces intra-frame motion. In retinal imaging, such a setup will provide better images because higher imaging speeds reduce motion artifacts.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.