Delineating 3D Angiogenic Sprouting in OCT Images via Multiple Active Contours Ting Xu 1 , Fengqiang Li 2 , Duc-Huy T. Nguyen 3 , Christopher S. Chen 3,4 , Chao Zhou 2 , and Xiaolei Huang 1 1 Department of Computer Science and Engineering, Lehigh University, Bethlehem, PA 18015, USA 2 Department of Electrical and Computer Engineering, Lehigh University, Bethlehem, PA 18015, USA 3 Department of Chemical and Biomolecular Engineering, University of Pennsylvania, Philadelphia PA 19104, USA 4 Department of Bioengineering, University of Pennsylvania, Philadelphia PA 19014, USA Abstract. Recent advances in Optical Coherence Tomography (OCT) has enabled high resolution imaging of three dimensional artificial vas- cular networks in vitro. Image segmentation can help quantify the mor- phological and topological properties of these curvilinear networks to facilitate quantitative study of the angiogenic process. Here we present a novel method to delineate the 3D artificial vascular networks imaged by spectral-domain OCT. Our method employs multiple Stretching Open Active Contours (SOACs) that evolve synergistically to retrieve both the morphology and topology of the underlying vascular networks. Quantifi- cation of the network properties can then be conducted based on the seg- mentation result. We demonstrate the potential of the proposed method by segmenting 3D artificial vasculature in simulated and real OCT im- ages. We provide junction locations and vessel lengths as examples for quantifying angiogenic sprouting of 3D artificial vasculature from OCT images. Keywords: Angiogenesis, Curvilinear Network, Active Contours, Optical Coherence Tomography. 1 Introduction Angiogenesis, a process where new vessels form from existing vasculature, re- lies on a series of highly coordinated events. During angiogenic sprouting, some of the most fundamental questions in vascular biology concern how endothelial cells coordinate to build branching tubular networks, which appear to be mor- phologically different in a tissue-specific manner. The molecular mechanisms to pattern tissue-specific vascular networks remain largely unknown. Previously, the angiogenic process has been observed, manipulated and stud- ied using confocal microscopy [1]. In this paper, we present an alternative ap- proach in which the 3D angiogenic process is monitored using OCT images. C.A. Linte et al. (Eds.): MIAR/AE-CAI 2013, LNCS 8090, pp. 231–240, 2013. c  Springer-Verlag Berlin Heidelberg 2013