Purchase this article with an account.
W.J. Donnelly, R.A. Applegate; Models of Human Cataract Validate Shack–Hartmann Nuclear Cataract Assessment . Invest. Ophthalmol. Vis. Sci. 2005;46(13):2912.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To use physical and computer models to suggest ways to improve metrics of nuclear cataract severity derived from Shack–Hartmann (S/H) images by optimizing exposure times and pupil size.
A physical model eye consisting of a fixed lens, variable pupil, cuvette, and retina was mounted to an S/H wavefront sensor. Located behind the pupil (retina side), the 5mm cuvette containing one of five polystyrene microsphere solutions served to vary nuclear cataract severity. The computer model was built in ZEMAX based on an exact replication of the physical eye model/wavefront sensor combination. We used these models to test the influence of exposure and pupil size on previously reported scatter metrics [JRS, 2004, 20(5), S515]. Exposure times were varied between 100ms and 200ms and pupil diameters were varied between 5mm and 7mm.
In both models, as cataract density increased light transmission decreased and scattered light increased. When pupil and exposure were allowed to vary, the forward scatter metric RMS_MEAN explained 59% of variance in simulated cataract severity. The coefficient of determination was improved to 64% by holding pupil size constant. When pupil size varied and exposure was held constant RMS_MEAN explained 61% of variance. By holding pupil diameter and exposure constant RMS_MEAN explained 71% of the variance in cataract severity.
The ability of S/H scatter metric RMS_MEAN to monitor changes in model cataract severity is significantly increased by holding pupil diameter and exposure time constant. The optimal exposure time should be selected to exploit the dynamic range of the camera without saturating the image and be fast enough to minimize eye motion blur.
This PDF is available to Subscribers Only