Purchase this article with an account.
S. Norrby, O. Findle, N. Hirnshcall, Y. Nishi, R. Bergman; Modelling the Pseudo-Phakic Eye for the Purpose of Sphero-Cylindrical IOL Power Calculation. Invest. Ophthalmol. Vis. Sci. 2010;51(13):5752.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To model the pseudo-phakic eye based on measurement to the extent possible and use to predict the sphero-cylindrical refractive outcome.
Preop axial length (AL), anterior chamber depth (ACD) were measured with IOLMaster. Postop ACD was measured with ACMaster. Pre- and postop corneal topography of both surfaces was measured with Pentacam, elevation data was extracted and fitted with Zernike polynomials. This information was used to model the pseudo-phakic eyes in Zemax to find the optimal sphero-cylindrical refraction by minimizing the wavefront aberration. This was compared to manifest refraction. Fifty patients were randomly implanted with an AMO TECNIS ZA9003 3-piece IOL in one eye and a ZCB00 1-piece IOL in the other. A prediction algorithm for postop ACD was established by partial least squares regression against preop AL and ACD. The estimated ACD was used to build predictive models of the postop eyes. Optimal refraction was calculated and compared to manifest. The sphero-cylindrical calculations used power vectors and tri-variate statistics was applied..
On postop data the mean calculated refraction was -0.61SD,-0.23CD@54°, while that found by manifest refraction was -0.27SD,-0.32CD@70°, the mean difference being +0.38SD,-0.17CD@93° calculated from the average power vector difference (P=10-5). The mean absolute error (MAE) was 0.42 D. The IOL position regression estimation had a standard deviation of 0.17 mm. A standard deviation of 0.30mm is typical in current IOL power formulas. The refraction calculated on preop data was -0.12SD,-0.13CD@132°, which differs on average from that measured by -0.15SD,-0.42CD@65° (P=10-8) with MAE 0.50 D.
he difference between postop refraction and that calculated based on postop measurements was small, which validates the approach, even though the difference was statistically significant. That not even better agreement was obtained is probably due to the large number of measured parameters involved. Every parameter has a random error and those add up by Gaussian error propagation. The MAE in SE is of the order of current cataract surgery.
This PDF is available to Subscribers Only