ESCRS - PP12.15 - Prediction Of Total Corneal Power From Anterior Surface Data In Keratoconus For Iol Calculation

Prediction Of Total Corneal Power From Anterior Surface Data In Keratoconus For Iol Calculation

Published 2022 - 40th Congress of the ESCRS

Reference: PP12.15 | Type: Free paper | DOI: 10.82333/b3eq-6p87

Authors: Umberto Camellin* 1 , Ivan Ninotta 2 , gianluigi Latino 1 , Pasquale Aragona 1 , Anna Maria Roszkowska 1

1Department of Biomedical and Dental Sciences and Morphofunctional Imaging, University of Messina, Italy,University of Messina,Messina,Italy, 2University of Messina,Messina,Italy

Purpose

Purpose: To evaluate different methods for prediction of the total corneal power in keratoconus using topographic parameters.

Setting

Setting: Cornea and Refractive Surgery Section, Ophthalmology Clinic, University Hospital of Messina, Italy.

Methods

Methods: Tomographic data from 280 eyes with keratoconus were evaluated. Anterior and posterior keratometry, anterior Q-value at 8mm, Central Corneal Thickness (CCT), location and value of Kmax, True Net Power (TNP) were measured with Scheimpflug’s camera Pentacam (Oculus, Germany). Calculated total corneal power (TCPc) at 3 mm was obtained using Gullstrand's formula. Predicted total corneal power at 3 mm (TCPp3) and at 4 mm (TCPp4) were obtained from univariate (TCPp3u, TCPp4u) and multivariate linear regression formulae (TCPp3m, TCPp4m). SimK, anterior Q-value, vertical location and value of Kmax were used in multivariate formulas. The mean absolute error (MAE) and median absolute error (MedAE) were calculated.

Results

Results: Highly significant correlations were found between: TCPp3u and TCPc (R2 = 0.978, p<.05), TCPp3m and TCPc(R2 = 0.989, p<.05). Lower but still significant correlations were found between: TCPp4u and TNP (R2 = 0.692, p<.05), TCPp4m and TNP (R2 = 0.887, p<.05).  Best results for TCP prediction at 3 and 4 mm were obtained with TCPp3m andTCPp4m: MAE of TCPp3m was 0.24 ± 0.20 (SD) D with MedAE of 0.20 D, MAE of TCPp4m was 0.96 ± 0.77 D with MedAE of 0.80 D.

Conclusions

Conclusions: Multivariate linear regression formulas using only anterior surface data can predict TCP with good approximation in eyes with keratoconus in case of absence of tomographic parameters.