Purchase this article with an account.
N. Doble, D. Miller, H. Zhao, G. Yoon, D.R. Williams; Deformable Mirror Requirements for Adaptive Correction of the Population of Normal Human Eyes . Invest. Ophthalmol. Vis. Sci. 2003;44(13):999.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: Adaptive optics (AO) provides the opportunity to increase the transverse and axial resolution of retinal imaging systems, as well as to correct the eye’s wave aberration in an automated phoropter. Our goal is to determine the specifications of various types of deformable mirror (DM) to correct the wave aberration in a population of normal eyes. Methods: At the Universities of Rochester and Indiana, two independent studies measured the ocular aberrations in a large population (109 and 70 eyes respectively) of human subjects. Using these data, several DM types were modeled to test their ability to correct these aberrations. The three DMs considered were a continuous face-sheet mirror, a segmented piston only (pure vertical motion per segment) mirror, and a segmented piston/tip/tilt mirror (3 degrees of freedom per segment). Their relative performance with varying numbers of actuators and stroke was modeled by calculating the Strehl ratio generated by the residual wave aberration. Results: For a 6.8mm pupil diameter, our simulations show that if defocus is corrected through the use of trial lenses then all three DM types require a stroke greater than 10 microns. Correcting astigmatism in addition reduces this to 6 microns. This is still greater than any DM available today. In order to achieve diffraction-limited correction, a continuous DM needs 13 actuators across a 6.8mm pupil, a piston only mirror would need 60 segments and the segmented piston/tip/tilt DM would require 11 segments. In addition, it is helpful if the mirror diameter is similar to the pupil diameter, i.e. 6-8mm, this minimizes the physical size of the whole AO system. The temporal demands of the DM are relatively light requiring a closed loop bandwidth of 1 Hz with a mirror update of 20-30 Hz. Conclusion: We have identified the parameters required to correct 95% of the population in our studies. The demand on the mirror stroke being the most difficult to achieve. Several DM designs currently being tested promise this high stroke coupled with sufficient actuators and small diameters. Meeting these DM requirements would allow retinal imaging and visual testing in a wide range of human subjects.
This PDF is available to Subscribers Only