An efficient method of evaluating the level of contrast of an OCT dataset is presented. The method develops a metric to segregate useful and not-so-useful data in one or more OCT B-scans, in order to reduce spurious subsequent analyses of the data by downstream segmentation algorithms. It is designed to be fast and efficient and is applied to determining autofocus of an OCT instrument real-time and in identifying a real image from its complex conjugate twin.
