TY - GEN
T1 - A methodology for analysis extraction and visualization of CT scans
AU - Eltonsy, N.
AU - Tourassi, G.
AU - Desoky, A.
AU - Elmaghraby, A.
N1 - Publisher Copyright:
© 2003 IEEE.
PY - 2003
Y1 - 2003
N2 - Compared to MRI computer tomography (CT) images have a very narrow signal attenuation range for soft tissues and very strong one for bones and air. The goal of this study is to design a simple reliable model to read, quantify, extract, and visualize the anatomical region of the liver from a colorectal CT scan. The work introduced is considered an important initial step for the development of computer assisted diagnosis (CAD) systems or for 3D reconstruction in realistic voxel-based rendering models. The methodology presented follows a two-stage approach. Initially, the angio-abdominal CT image is carefully analyzed to amplify the CT soft tissues' signals. The task is achieved by optimizing the threshold values for 2-D visualization, background discrimination, and identification of the CT slice with the largest liver bulk. Consequently, a technique is proposed to granulate the images on a per slice basis. The intensity based granulation technique is set to 9.9 urn similarity difference and supported by strongly connected image map created from extracted features with 98% neighborhood ratio' threshold. The proposed two-step methodology was successfully tested on 181 colorectal CT scans.
AB - Compared to MRI computer tomography (CT) images have a very narrow signal attenuation range for soft tissues and very strong one for bones and air. The goal of this study is to design a simple reliable model to read, quantify, extract, and visualize the anatomical region of the liver from a colorectal CT scan. The work introduced is considered an important initial step for the development of computer assisted diagnosis (CAD) systems or for 3D reconstruction in realistic voxel-based rendering models. The methodology presented follows a two-stage approach. Initially, the angio-abdominal CT image is carefully analyzed to amplify the CT soft tissues' signals. The task is achieved by optimizing the threshold values for 2-D visualization, background discrimination, and identification of the CT slice with the largest liver bulk. Consequently, a technique is proposed to granulate the images on a per slice basis. The intensity based granulation technique is set to 9.9 urn similarity difference and supported by strongly connected image map created from extracted features with 98% neighborhood ratio' threshold. The proposed two-step methodology was successfully tested on 181 colorectal CT scans.
KW - Attenuation
KW - Biological tissues
KW - Bones
KW - Computed tomography
KW - Computer aided diagnosis
KW - Design automation
KW - Image reconstruction
KW - Liver
KW - Magnetic resonance imaging
KW - Visualization
UR - http://www.scopus.com/inward/record.url?scp=84948143433&partnerID=8YFLogxK
U2 - 10.1109/ISSPIT.2003.1341162
DO - 10.1109/ISSPIT.2003.1341162
M3 - Conference contribution
AN - SCOPUS:84948143433
T3 - Proceedings of the 3rd IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2003
SP - 479
EP - 482
BT - Proceedings of the 3rd IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2003
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 3rd IEEE International Symposium on Signal Processing and Information Technology, ISSPIT 2003
Y2 - 14 December 2003 through 17 December 2003
ER -