considered the gold standard tech- nique for the ... ule and ultrasound pachymetry (USP), and to assess repeatability of ..... may be related to the high degree of.
Acta Ophthalmologica 2012
Corneal thickness mapping by 3D swept-source anterior segment optical coherence tomography Alberto Neri, Marianna Malori, Patrizia Scaroni, Rosachiara Leaci, Elisabetta Delfini and Claudio Macaluso Ophthalmology, University Hospital of Parma, Parma, Italy
ABSTRACT. Purpose: To assess accuracy and repeatability of central corneal thickness (CCT) measurements obtained by swept-source anterior segment optical coherence tomography (AS-OCT), spectral-domain retinal OCT with corneal module and ultrasound pachymetry (USP), and to assess repeatability of pachymetric mapping with AS-OCT. Methods: 50 healthy volunteers were recruited. A single, experienced operator analysed the right eye of each participant twice in the same session with ASOCT (‘corneal map’ routine), retinal OCT and USP. CCT measurements were compared using repeated-measures analysis of variance, Bonferroni test, Pearson correlation and Bland-Altman plots. Repeatability of thickness maps and CCT measurements were assessed using Alpha of Cronbach, intraclass correlation coefficient (ICC) and coefficient of repeatability. Results: Mean CCT ± SD was 540 ± 28.9 lm for AS-OCT, 544 ± 29.5 lm for retinal OCT and 549.3 ± 31.7 lm for USP; the differences were statistically significant (p < 0.01). CCT measurements obtained with the three instruments were highly correlated: r was 0.965 for ASOCT ⁄ USP, 0.962 for retinal OCT ⁄ USP and 0.984 for AS-OCT ⁄ retinal OCT comparison. The repeatability of CCT measurements was higher for AS-OCT than for the other devices (p < 0.001). Repeatability of pachymetric maps was excellent (ICC = 0.999). Conclusions: Pachymetric maps by swept-source AS-OCT showed excellent repeatability. CCT measurements obtained by AS-OCT, USP and retinal OCT were highly correlated although not identical. Key words: agreement – anterior-segment OCT – corneal thickness – pachymetric map – repeatability – swept-source OCT
Acta Ophthalmol. 2012: 90: e452–e457 ª 2012 The Authors Acta Ophthalmologica ª 2012 Acta Ophthalmologica Scandinavica Foundation
doi: 10.1111/j.1755-3768.2012.02453.x
Introduction Pachymetry, the measurement of corneal thickness, is part of the clinical rou-
e452
tine in the study of corneal anatomy and pathology, in the evaluation of glaucoma risk and in the preliminary phases of corneal and refractive surgery.
Ultrasound pachymetry (USP) is considered the gold standard technique for the clinical evaluation of corneal thickness, and it has proven good reliability and repeatability. The main limit of USP is its invasiveness, as it requires the ultrasound probe to contact the eye causing patient’s discomfort, risk of infection and of epithelial damage. This also decreases the utility of USP in paediatric ophthalmology and immediately after surgery. Moreover, USP is partially operatordependent as it may be subject to errors deriving from the deformation of the cornea, in case of excessive pressure during the examination, or from decentred or oblique positioning of the probe on the corneal surface. Several noninvasive techniques to measure corneal thickness are available, main options being specular microscopy, very high-frequency digital ultrasound, scanning slit tomography, Scheimpflug camera and optical coherence tomography (OCT) devices. OCT instruments allow in vivo, noncontact, high-definition visualization and measurement of the eye structures. OCT systems obtain corneal thickness values from the analysis of sectional images of the anterior segment of the eye, after identification of the external and internal boundaries of the cornea. Corneal thickness measurements can be obtained both by retinal OCTs equipped with anterior segment modules or by dedicated anterior segment OCTs (AS-OCT), which can visualize the entire anterior segment.
Acta Ophthalmologica 2012
In the last years, OCTs have become widely used in the study of corneal and irideal pathologies, cataract and glaucoma. They can also perform a topographical analysis of the examined corneal tissue, obtaining pachymetric maps (Li et al. 2006; Mohamed et al. 2007; Li et al. 2008b; Li et al. 2010). Recently, spectral-domain (SD) ASOCTs devices have been developed, showing improved resolution and higher acquisition rates, in the same way as it happened when SD retinal OCTs succeeded to time-domain devices. Faster scan acquisition rates, in particular, are expected to reduce the artefacts because of eye movements that were the principal limitation for time-domain AS-OCT to obtain topographical analysis of the cornea, as maps of both corneal curvature (Tang et al. 2006) and thickness (Li et al. 2006, 2008a,b, 2010). Swept-source OCT is a particular form of spectraldomain OCT technology characterized by high-speed acquisition (Yun et al. 2003). A swept-source OCT prototype allowing 3D imaging and biometry of the anterior segment of the eye was studied by Fukuda et al. (Fukuda et al. 2009, 2010). Herein, we studied the accuracy and repeatability of automated central corneal thickness (CCT) measurements obtained by a commercially available swept-source AS-OCT. CCT measurements obtained with AS-OCT were compared with those obtained with USP and retinal SD-OCT with corneal module. Moreover, the repeatability of pachymetric maps by sweptsource AS-OCT was analysed.
Methods All the participants in the study provided informed consent and were treated in accordance with the principles of the Declaration of Helsinki. The right eyes of fifty volunteers, with no ophthalmic pathologies except mild refractive errors, were included in the study. A single, experienced operator analysed all the included eyes with three different devices: (i) a swept-source AS-OCT (CASIA-OCT, produced by Tomey Co., Nishi-Ku, Nagoya, Japan), (ii) a retinal OCT with corneal module (Cirrus-OCT, produced by Zeiss Inc., Oberkochen, Germany) and (iii) an ultrasound pachymeter (AL-2000, produced by Tomey Co., Japan).
All the measurements on each eye were performed twice in a single session (maximum overall time 2 hr) to analyse the repeatability of corneal thickness measurements. Central corneal thickness measurements and pachymetric maps by AS-OCT
CASIA-OCT (produced by Tomey Co.) is a commercially available SD AS-OCT using a swept-source laser emitting light with wavelength centred at 1310 nm. Its main features are axial and transverse resolution, respectively, of 10 and 30 lm, scanning speed of 30 000 A-scans ⁄ sec and a scan pattern that is widely customizable into a maximal scanning width of 16 · 16 mm and a maximum depth of 6 mm in tissue. The AS-OCT is equipped with an automated corneal topography routine, which was used in this study for the acquisition of corneal scans. The scan pattern consisted of 16 radial B-scan of 10 · 4mm (512 A-scan for each line), centred on the corneal vertex (fixation axis); centring during the scan acquisition was automatically performed by the device, while the subject was fixating an internal aim. The acquisition time of the corneal topography routine is
