Abstract
Purpose: :
To describe an image analysis method and software for quantitatively analyzing corneal images obtained using an ultra-high resolution optical coherence tomography (OCT) and to assess its validity.
Methods: :
A custom built spectral-domain OCT with ultra-high resolution (~3 µm) was built for imaging the anterior segment. All images were obtained in 20 healthy subjects. Image processing steps applied to the ultra-high resolution OCT images that were constructed by 2048x1365 pixels. First, the surface between air and anterior cornea was semi-automatically segmented by 3 to 4 points on the interface and fitted in a spline polynomial model. Then the OCT image was processed to correct the distortion due to the light refraction at the air- cornea interface based on the Fermat’s principle. From the corrected image, the algorithm semi-automatically defined the boundaries for cornea and its sub-layers by user input of 3 to 4 points. The points were identified by the signal peaks at the interface. The spline polynomial model was also applied to fit the contour based on the selected points. The corneal thickness was measured as the distance between the anterior and posterior corneal surfaces along lines perpendicular to the anterior surface at the point of measurement. Along these lines, the epithelium thickness was also measured as the distance between the anterior corneal surface and basal cell layer. A profile for corneal and epithelium thickness was generated from each meridional cross section. Sixteen corneal cross sections were loaded to compute the corneal and epithelium thickness map by interpolation.
Results: :
The analysis software successfully measured all parameters in all images and dimensional results from the obtained OCT images. The corneal and epithelial thickness maps were successfully created and consistent with previous literature. The segmented boundaries agreed well with visual inspection of each cross-sectional image.
Conclusions: :
We have developed image analysis method and software for analyzing ultra-high resolution OCT images to quantitatively measure corneal and epithelial thicknesses.
Keywords: imaging/image analysis: non-clinical • image processing • anterior segment