ABSTRACT
In virtual colonoscopy (VC) approaches, cameras are moved along the centerline of a colon to visualise its inner surface. The orientation of each camera depends on its position and the optical axis (look-at) and view up vectors control the camera direction. However, the look-at and view-up vectors are changing at each position along the centerline due to bending and torsion in the colon surface. Therefore, this rough camera movement generates an unsmooth image sequence (i.e., large differences between consecutive frames), which is not comfortable to the observer’s eyes. To solve this problem, we propose a transformation method that embeds the 3D colon centerline into a 2D plane or a 1D line. We can use this transformation to stabilise the camera movement in different VC visualisation methods. We use the visualisation loss to quantify the feasibility of the proposed method by comparing the visualisation quality before and after applying the proposed transformation.
Acknowledgement
This Work has been supported by Kentucky Imaging Technologies, The National Science Foundation (Project:1602333) and the National Institutes of Health (Project: 1R43CA250750 - 01).
Disclosure statement
No potential conflict of interest was reported by the author(s).
Additional information
Notes on contributors
Mostafa Mohamed
Mostafa Mohamed received the M.S. degree in Computer engineering from Cairo University, Egypt in 2014. Currently he is a Ph.D candidate in the Speed School of Engineering at University of Louisville. He is also a member in the Computer Vision and Image Processing (CVIP) Lab since 2015.
Asem Ali
Asem M. Ali received the M.S. degree in electrical engineering from Assiut University, Asyut, Egypt, in 2002, and the Ph.D. degree in computer engineering from the University of Louisville, Louisville, KY, USA. He is currently a Research Scientist with the Computer Vision and Image Processing Laboratory. His research interests include image analysis, machine learning, face recognition, and facial expressions and emotions recognition. He has authored over 40 papers in journals and conferences.
Salwa Elshazly
Salwa Elshazly, holds a Bachelor of Science degree in Biology from Ain-Shams University, Egypt, and a postgraduate Computer Science degree from Jefferson Community College (JCC), Louisville, KY. She was the top-ranked graduate of her Class. From 1997-2013, she was a biomedical researcher at the School of Engineering, University of Louisville (UofL), where she conducted research on biomedical data analysis. She has co-authored over 40 scientific papers and technical reports on the detection of colon and lung cancers using low dose CT imaging. She is the founder and CEO of Kentucky Imaging Technologies (www.kyimaging.com) which is developing technologies for visualization of CT Colonography (CTC), Chest CT and Imaging of the Oral Cavity. She has been PI of an NIH grant and three grants from the KY Economic Development Cabinet.
Aly Farag
Aly Farag, Fellow, IEEE and IAPR: received B.S. in EE from Cairo Univ. M.S. in Bioengineering from the Ohio State and the Univ. of Michigan, and PhD in EE from Purdue. He is a Professor of ECE at the Univ. of Louisville, and director of the Computer Vision & Image Processing Laboratory, focusing on research and teaching in computer vision, biometrics and biomedical imaging. He introduced over 13 new courses into the ECE curriculum, authored over 400 papers, edited two volumes on deformable models and a textbook on Biomedical Image Analysis (Cambridge Univ. Press, 2014). He graduated over 70 MS and PhD students, and mentored over 20 postdoctoral researchers. He holds seven US patents on object modeling, computer-aided diagnosis, and visualization. He was lead editor of IEEE-TIFS special issue on Face Recognition in the Wild (December 2014), and co-general chair of ICIP-2009. He is recipient of the University top Awards: Research (1999), Teaching (2009, 2011) and Trustees (2015).