Abstract
Purpose: :
It is often observed that ocular aberrations undergo dynamic variations over short periods of time. We propose to quantify and model these aberrations based on appropriate measurements. This work may be helpful in understanding the physiological causes of ocular aberration dynamics.
Methods: :
A Shack-Hartmann wavefront sensor has been built and calibrated to meet our needs. It consists of a very sensitive camera that records data at 37.2 Hz, coupled with a lenslet array that samples the exit pupil of the eye every 250 microns. A scanning mirror is used to average down to zero unwanted effects of speckle. We measured young healthy subjects under steady-state experimental conditions. Trials were typically 30 seconds long, including 5 instances of blinking.
Results: :
Wavefronts were reconstructed using a Zernike polynomial expansion. Time-frequency plots of the Zernike modes showed that the ocular aberration dynamics exhibit non-stationary behaviour, which complicated the modeling procedure. An Autoregressive Integrated Moving Average (ARIMA) modeling approach was employed to model the Zernike mode time series. The correlations between different Zernike modes were also considered. The completed models were used to generate realistic simulated data that could be treated as sample realisations of the true process.
Conclusions: :
The ARIMA approach provided accurate models and proved to be an effective tool for generating simulated aberration data. These tools will be applied to a larger population of subjects, and will hopefully be of use in the design and testing of retinal imaging systems.
Keywords: optical properties • aberrations • computational modeling