Robotically assisted large-field-of-view, ultra-high-resolution imaging for the entire cornea with spectral domain optical coherence tomography

It is challenging to achieve high-resolution, high signal-to-noise ratio (SNR) imaging of the entire cornea with optical coherence tomography (OCT) due to the cornea's curved, dome-like shape. With a typical telecentric scanning optical path, the scattered photons are less likely to return to the interferometer at the peripheral regions, leading to a reduction in SNR. This study presents a 6-degree-of-freedom (6-DOF) robotic-arm-assisted, ultra-high-resolution OCT imaging system designed to adapt to the cornea's curvature. In this system, a multi-region scanning approach combined with robotic arm pose control is employed to maintain the sample beam perpendicular to the surface across the entire cornea for consistent high SNR imaging. An en face-correlation-based volumetric registration algorithm is proposed for high-accuracy registration for the obtained multiple local images. Experiments on <i>ex vivo</i> porcine eyes demonstrated that the proposed technique achieved high-quality imaging over a 12 mm × 12.4 mm area, covering the cornea and its peripheral regions. This system clearly visualized structures such as the epithelial, stromal, and endothelial layers, validating its potential for comprehensive diagnosis and evaluation of corneal diseases. Overall, we believe this study could provide a solution for high-resolution imaging on cornea in ophthalmology.