Purchase this article with an account.
Christy K Sheehy, Ramkumar Sabesan, Pavan N Tiruveedhula, Qiang Yang, Austin Roorda; Active eye-tracking for AOSLO. Invest. Ophthalmol. Vis. Sci. 2014;55(13):5195.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To demonstrate a hybrid tracking scanning laser ophthalmoscope (TSLO) and adaptive optics scanning laser ophthalmoscope (AOSLO) system with both optical and digital eye-tracking capabilities.
A TSLO was optically combined, via dichroic beamsplitter, with an AOSLO to provide active hardware-based eye-tracking capabilities. The TSLO system uses 730 nm illumination and the AOSLO uses 840 nm. Retinal image-based eye-tracking was performed over a 3.5° FOV in real-time with the TSLO system. Eye motion was reported at a rate of 960 Hz and the x and y displacements of eye motion were converted into two 14-bit voltage outputs, then low-pass filtered to remove residual high-frequency noise. The outputs were used to drive a tip/tilt mirror conjugate to the pupil plane of the AOSLO system, which actively moved to keep the AOSLO imaging raster on target. Remaining high spatial resolution eye motion artifacts in the smaller FOV AOSLO system were corrected using previously reported image-based software methods. System performance was quantified using both model and human eyes. The model eye was equipped with a galvo scanner that allowed for control of the frequency and amplitude of simulated eye motion in both x and y axes. Input frequencies and amplitudes were chosen at values where AOSLO image-based software tracking alone would fail.
The hybrid system provided active steering of the tip/tilt mirror of the AOSLO system for real-time stabilized imaging. Using the model eye, the AOSLO videos, prior to digital stabilization, showed a >90% reduction in amplitude for frequencies lower than 2.3 Hz and >65 % reduction for those below 10.3 Hz. For the human eye, we looked at four 10 second videos in both systems. On average, for frequencies lower than 3 Hz, there was up to a 14 times reduction in amplitude for y (6 times in x) between the original motion in TSLO videos and stabilized AOSLO videos with active tracking. Further removal of motion using real-time digital correction reduced the residual motion in the videos to previously reported levels.
By correcting for high amplitude, low frequency drifts of the eye, the active TSLO eye-tracking system enables the AOSLO to successfully capture high-resolution retinal images over a larger range of motion than previously possible. Having a larger FOV TSLO is beneficial for imaging subjects and patients with large fixational eye movements that would otherwise cause image-based tracking methods to fail.
This PDF is available to Subscribers Only