Recently, time-domain full-field OCT (TD-FF-OCT) evolved into the tool for in vivo eye exploration. Particular benefits of TD-FF-OCT include cellular resolution, rapid en face imaging without scanning artifacts, reduced sensitivity to optical aberrations, and non-contact performance. One limitation of TD-FF-OCT is related to its ’time-domain’ nature – tomographic image is reconstructed from several consecutive (in time) camera frames. The sample should stay static, otherwise the signal is reduced. In this work we present TD-FF-OCT that can capture in vivo human corneal images in a single camera shot. This became possible by using a combination of optical/image processing and deep learning techniques.
|