Purchase this article with an account.
B. Lamory, B. Sahin, F. Harms, S. Berthier; Performance Assessment of a Pupil Tracking System for Adaptive Optics. Invest. Ophthalmol. Vis. Sci. 2008;49(13):4199.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Adaptive Optics (AO) is particularly suitable for correction of aberrations that change over time - a necessity for high resolution imaging of the retina. However, AO imaging performance is effected by eye motion and so wavefront sensors with high repetition rates are often required. We developed a new approach for enhancing aberration correction in AO retinal imaging by integrating a Pupil Tracking System (PTS) into the AO loop. In this study we assessed the performance of the PTS developed for this purpose.
The PTS functions using digital image processing of the data acquired by a high resolution CCD camera over a 13 X 5 mm2 imaging area while the pupil is illuminated by a LED (NIR) array. Real-time image processing algorithms that include thresholding, pupil edge detection and parabolic fitting were developed.First, we established a baseline standard by testing the system’s performance using an artificial eye with a pupil diameter of 7 mm, mounted on a two-axis micrometric motion controller system. Next, the PTS’ performance was tested in vivo through series of 1024 consecutive pupil diameter measurements on five subjects after their pupils were paralyzed by applying mydriatic drops. The subjects’ heads were stabilized with a standard ophthalmic chinrest.
Real-time pupil detection was achieved for movements on both the vertical and horizontal axes over a range of ±2 mm. The characterization of the PTS’ performance showed a temporal resolution of 90 Hz, an accuracy of ±17 µm with a standard deviation of ±6.2 µm over a ±2 mm range of movement, and an accuracy of ±6 µm with a deviation of ±1.5 µm over ±1 mm range, both horizontally and vertically. In vivo experimental results demonstrated a standard deviation of ±19 µm.
This system effectively tracked minute pupil movements with higher precision when compared to that of known existing eye tracking systems. Its high resolution, over ±2 mm range of movements, makes it suitable for integration into the AO loop for retinal imaging. This will reduce the AO system’s dependence on the wavefront sensors’ speed, enabling the use of conventional components for the construction of the final AO retinal imaging system.
This PDF is available to Subscribers Only