Computational adaptive optics for optical coherence tomography using multiple randomized subaperture correlations

Computational adaptive optics (CAO) is emerging as a viable alternative to hardware-based adaptive optics—in particular when applied to optical coherence tomography of the retina. For this technique, algorithms are required that detect wavefront errors precisely and quickly. Here we propose an extension of the frequently used subaperture image correlation. By applying this algorithm iteratively and, more importantly, comparing each subaperture not to the central subaperture but to several randomly selected apertures, we improved aberration correction. Since these modifications only slightly increase the run time of the correction, we believe this method can become the algorithm of choice for many CAO applications.