Image Quality Comparison Among Spectral Domain Optical Coherence Tomography Systems

Abstract

Purpose: : Multiple spectral domain optical coherence tomography (SD-OCT) options are available to image the retina clinically. Accurate structural information provides valuable information toward patient management. We developed a method to compare image quality among commercially available SD-OCT imaging in normal eyes and those with macular disease.

Methods: : OCT images (N= 200) in the vertical or horizontal meridians across the foveal center were obtained in patients with clear media in normal eyes (N=20) and those with macular disease (N=5) using five commercially available spectral domain systems: Cirrus HD-OCT, Topcon 3D OCT-1000, Heidelberg Engineering, Opko/OTI Instrumentation Spectral OCT SLO, and Optovue. All images for each patient were obtained within one hour following pupillary dilation.Image quality assessment was performed using masked evaluators. Subjective overall image quality evaluation was performed for each eye by rank ordering OCT images from each of 5 SD-OCT systems. Additionally, 5 categories of inner and outer retinal structures were assessed: identification of the posterior hyaloid (PH); distinction between the ganglion cell layer (GCL) and nerve fiber layer; between inner nuclear layer, outer plexiform layer, and outer nuclear layer; between interdigitating components of the outer segment and retinal pigment epithelium (RPE); and identification of the posterior margin of the choroidal vessels.An objective image quality methodology was developed and performed using Gaussian smoothing until the following structures were indistinguishable: inner-outer segment junction, external limiting membrane, RPE, and PH.

Results: : All 5 SD-OCT systems provided similarly distinct imaging of the middle retinal layers. The cycles of Gaussian smoothing to achieve this endpoint was recorded for all images (p>0.2). However, two of these systems enabled consistent visualization of the posterior hyaloid and cortical vitreous, GCL, and differentiation of the outer segment/RPE complex. Increased image quality was significant for both subjective and objective evaluation (p < 0.05).

Conclusions: : Distinction between SD-OCT systems may lie in better visualization of the inner- and outer-most retinal layers. Macular disease specific to these retinal layers maybe be better elucidated using these OCT systems.

Keywords: imaging/image analysis: clinical • retina • imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound)