TY - JOUR
T1 - Spatial-temporal image-constrained lung 4D-CT reconstruction for radiotherapy planning
AU - He, Tiancheng
AU - Xue, Zhong
AU - Yu, Nam
AU - Teh, Bin S.
AU - Wong, Stephen T.
N1 - Publisher Copyright:
© Springer International Publishing Switzerland 2014.
Copyright:
Copyright 2015 Elsevier B.V., All rights reserved.
PY - 2014
Y1 - 2014
N2 - Thoracic radiotherapy planning is increasingly dependent on 4D computed tomography (CT), which acquires axial images in multiple respirator phases and reconstructs them into 3D CT images based on respiratory signals. However, large reconstruction errors or artifacts may be observed due to poor reproducibility of breathing cycles. In this paper, 4D-CT reconstruction of helical mode CT scanning is achieved by incorporating spatial continuity and longitudinal smoothness of anatomical structures, such as chest surface, bone, vessel, and lung fields. The objective is to optimize the assignment of each axial image into different respiratory phases so that the artifacts or spatial discontinuity of anatomical structures are minimized, and the anatomical structures maintain their longitudinal consistency. In experiments, we compared our results visually and quantitatively with the current surrogate-based, image-matchingbased, and chest surface-constrained methods. The results showed that the proposed algorithm yields better helical mode 4D-CT than other proposed methods
AB - Thoracic radiotherapy planning is increasingly dependent on 4D computed tomography (CT), which acquires axial images in multiple respirator phases and reconstructs them into 3D CT images based on respiratory signals. However, large reconstruction errors or artifacts may be observed due to poor reproducibility of breathing cycles. In this paper, 4D-CT reconstruction of helical mode CT scanning is achieved by incorporating spatial continuity and longitudinal smoothness of anatomical structures, such as chest surface, bone, vessel, and lung fields. The objective is to optimize the assignment of each axial image into different respiratory phases so that the artifacts or spatial discontinuity of anatomical structures are minimized, and the anatomical structures maintain their longitudinal consistency. In experiments, we compared our results visually and quantitatively with the current surrogate-based, image-matchingbased, and chest surface-constrained methods. The results showed that the proposed algorithm yields better helical mode 4D-CT than other proposed methods
KW - 4D-CT reconstruction
KW - Bayesian model
KW - Registration
KW - Respiratory motion
UR - http://www.scopus.com/inward/record.url?scp=84921499653&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84921499653&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-13909-8_16
DO - 10.1007/978-3-319-13909-8_16
M3 - Article
AN - SCOPUS:84921499653
VL - 8680
SP - 126
EP - 133
JO - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
JF - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SN - 0302-9743
ER -