Accuracy and repeatability of net corneal astigmatism measured with a novel OCT topography algorithm compared to two commercial devices

Abstract

Purpose : To assess the accuracy of net corneal astigmatism (NCA) measured with a spectral-domain optical coherence tomography (SD-OCT, Avanti, Optovue Inc), a Scheimpflug imaging device (Pentacam HR, Oculus Optikgerate GmbH) and a swept-source optical biometer (IOLMaster 700, Carl Zeiss Meditec AG) and to determine the repeatability for Avanti and Pentacam HR derived astigmatism.

Methods : Sixty pseudophakic eyes from 39 subjects (39 eyes post-LASIK, 16 eyes post-radial keratectomy and 5 eyes post-photorefractive keratectomy) with monofocal non-toric intraocular lens (IOL) were measured with Avanti, Pentacam HR and IOLMaster 700. Avanti topography was measured using a 16-meridian scan pattern that was repeated 10 times in less than 1.50 sec. Motion error was reduced by comparison of the 10 redundant scan sets and NCA was obtained by Zernike decomposition of the vertex-centered 4 mm diameter area of anterior and posterior corneal surfaces using a custom algorithm. For Pentacam HR and IOLMaster 700, true net and total keratometry readings were used, respectively. Vector analysis was used by decomposing astigmatism into cardinal and oblique components. Repeatability was assessed by means of the coefficient of repeatability (CR = S_w * 2.77, where S_w is the within-subject standard deviation of three repeated measures). To assess accuracy, NCA was compared to manifest refraction (MR) converted to the corneal plane. Vectors difference between NCA and MR were calculated and the average of the absolute magnitude of the difference vectors (mean absolute difference) and the mean difference vector reported.

Results : The repeatability of NCA measurement was significantly better (F test, p<0.01) for Avanti (total vector CR = 0.29 D) than for Pentacam HR (total vector CR = 0.67 D). The accuracy of NCA measurement was best for Avanti compared to Pentacam HR and IOLMaster 700 (Fig. 1) – the mean absolute difference with MR was significantly lower (p<0.01 by mixed effect model using gamma distribution).

Conclusions : In post refractive surgery patients, NCA measured by high-speed SD-OCT can achieve better accuracy than current commercial Scheimpflug and swept-sourct biometer devices. This may be useful in toric IOL selection.

This is a 2021 ARVO Annual Meeting abstract.

 

Figure 1. Double-angle plots of the difference vectors between manifest refraction and corneal astigmatism.