Correcting intra-volume distortion for AO-OCT using 3D correlation based registration

Abstract

Adaptive optics (AO) based ophthalmic imagers, such as scanning laser ophthalmoscopes (SLO) and optical coherence tomography (OCT), are used to evaluate the structure and function of the retina with high contrast and resolution. Fixational eye movements during a raster-scanned image acquisition lead to intra-frame and intra-volume distortion, resulting in an inaccurate reproduction of the underlying retinal structure. For three-dimensional (3D) AO-OCT, segmentation-based and 3D correlation based registration methods have been applied to correct eye motion and achieve a high signal-to-noise ratio registered volume. This involves first selecting a reference volume, either manually or automatically, and registering the image/volume stream against the reference using correlation methods. However, even within the chosen reference volume, involuntary eye motion persists and affects the accuracy with which the 3D retinal structure is finally rendered. In this article, we introduced reference volume distortion correction for AO-OCT using 3D correlation based registration and demonstrate a significant improvement in registration performance via a few metrics. Conceptually, the general paradigm follows that developed previously for intra-frame distortion correction for 2D raster-scanned images, as in an AOSLO, but extended here across all three spatial dimensions via 3D correlation analyses. We performed a frequency analysis of eye motion traces before and after intra-volume correction and revealed how periodic artifacts in eye motion estimates are effectively reduced upon correction. Further, we quantified how the intra-volume distortions and periodic artifacts in the eye motion traces, in general, decrease with increasing AO-OCT acquisition speed. Overall, 3D correlation based registration with intra-volume correction significantly improved the visualization of retinal structure and estimation of fixational eye movements.

Effect of fixational eye movements in corneal topography measurements with optical coherence tomography

“…Anderson et al [14] obtained data from different points of the cornea simultaneously using a multichannel acquisition OCT and reported a repeatability on the order of 0.1 D. Other scan types such as a spiral with isotropic transverse sampling [31], Lissajous curves [32] and orthogonal raster scan patterns [33] have been proposed for retinal imaging OCT, where features in the images can help to register the data with each other. Also, using a classical raster scan pattern there are methods to correct intra-volume motion correction algorithms [34], but to our knowledge, the applicability of these methods to calculate the topography of the cornea and the crystalline lens remains unexplored. The development of models to accurately simulate data acquisition, such as the one presented here, could clarify the improvement in the quality of the data obtained with each of them and the variability dependence with scan parameters.…”

Shot-noise limited phase-sensitive imaging of moving samples by phase-restoring subpixel motion correction in Fourier-domain optical coherence tomography

“…For repeated volumetric C-scans in human ORG experiment, we followed the coarse-to-fine strategy proposed by Do et al to estimate the translational motion of individual B-scans [34, 35], except that in the fine estimation step, we adopted single-step DFT approach to extend the previous pixel-level estimation to subpixel-level in the depth ( z ) and the fast-scan ( x ) dimensions. Briefly, in the coarse estimation step, a set of target sub-volumes consisting of several consecutive B-scans were selected from the target volume, and their positions in the reference volume were estimated with threedimensional (3D) NCC method.…”

“…For high-speed phase-sensitive OCT imaging with repeated cross-sectional B-scans or repeated volumetric C-scans, image registration must be conducted in 2 or 3 DoFs in both the axial and the lateral directions. In previous studies, translational displacements between the reference and target images were estimated using normalized cross-correlation (NCC) based [28][29][30][31] or phase-only correlation (POC) based methods [32], while the image correction was achieved by shifting the complex-valued target image [33][34][35] or the 2~4 fold upsampled target image [28,29] in the unit of pixels. However, these techniques can only shift the OCT image by discrete distances restricted by the upsampling rate, and the computational and storage load will increase dramatically with more refined displacements.…”

Characterizing Cone Spectral Classification by Optoretinography

“…Each recorded 2D spectrum ( I ( x, y, λ )) was treated to background subtraction, k-space resampling ( I ( x, y, k )), and a Fourier transform I ( x, y, z ) to yield an OCT volume. All volumes were registered using segmentation-based 3D registration[24]. Once the volumes were registered, each OCT volume was referenced to the mean of all the volumes that were recorded before the start of the stimulus to cancel the arbitrary phase at each pixel.…”