Abstract
Purpose :
We recently introduced a method to obtain single-pass transmission images of the fundus. The method exploits the relatively deep penetration of near-infrared (NIR) light in tissue to transcranially illuminate the fundus. In this paper, we compare images obtained with the transcranial method to those obtained quasi-simultaneously with conventional transpupil fundus reflection photography.
Methods :
For transcranial fundus illumination, we coupled NIR light from a high-power LED into a fiber bundle and placed the distal end on the subject’s temple. For reflection imaging, we routed light from a separate LED (at the same wavelength) into the illumination path of a commercial non-mydriatic fundus camera. The fundus camera was equipped with a sCMOS camera. By toggling the two LED channels, in synchrony with the free running camera, we obtained simultaneous reflection and transmission fundus images. 64 frames, corresponding to 2-3 cardiac cycles, were registered, averaged, and flattened for display to improve contrast to noise.
Results :
Example images of fundus reflection (left column) and transmission (right column) are shown in the attached figure. Notably, in the transmission image, the optic disc is bright compared with its surroundings, likely due to the lack of pigmented epithelium at this location. Also in transmission, the large retinal vessels (shown in the second row zoomed areas) are uniformly dark across vessel lumen, in contrast with the reflection images, which exhibit varying levels of central reflex.
Conclusions :
NIR light penetrates deeply enough through the skull and head to enable diffuse transmission fundus imaging. This illumination geometry has advantages, including overall simplicity: the entire naturally-dilated pupil may be used to collect images. The transmission geometry is also advantageous for fundus spectroscopy, in contrast with fundus reflectometry where the presence of several light paths (from multiple reflections) adds significant complexity to light transport models.
This abstract was presented at the 2019 ARVO Imaging in the Eye Conference, held in Vancouver, Canada, April 26-27, 2019.