TY - JOUR AU - Gawish, Ahmed AU - Haines, L. AU - Marschall, S. AU - Wong, Alexander AU - Sorbara, L. AU - Bizheva, Kostadinka AU - Fieguth, Paul PY - 2016/10/03 Y2 - 2024/03/29 TI - Improved OCT Human Corneal segmentation Using Bayesian Residual Transform JF - Journal of Computational Vision and Imaging Systems JA - J. Comp. Vis. Imag. Sys. VL - 2 IS - 1 SE - Articles DO - 10.15353/vsnl.v2i1.117 UR - https://openjournals.uwaterloo.ca/index.php/vsl/article/view/117 SP - AB - <p>The inherent poor signal to noise ratio of Optical Coherent Tomography<br />(OCT) is considered as a main limitation of OCT segmentation,<br />particularly because images are sampled quickly, at high resolutions,<br />and in-vivo. Furthermore, speckle noise is generated by<br />the reflections of the OCT LASER limits the ability of automatically<br />segmenting OCT images. This paper presents a novel method to<br />automatically segment human corneal OCT images. The proposed<br />method uses Bayesian Residual Transform (BRT) to build a noise<br />robust external force map, that guides active contours model to the<br />corneal data in OCT images. Experimental results show that the<br />proposed method outperforms the classical as well as the state-ofthe-<br />art methods.</p> ER -