%0 Journal Article
%T Improving Intensity-Based Lung CT Registration Accuracy Utilizing Vascular Information
%A Kunlin Cao
%A Kai Ding
%A Joseph M. Reinhardt
%A Gary E. Christensen
%J International Journal of Biomedical Imaging
%D 2012
%I Hindawi Publishing Corporation
%R 10.1155/2012/285136
%X Accurate pulmonary image registration is a challenging problem when the lungs have a deformation with large distance. In this work, we present a nonrigid volumetric registration algorithm to track lung motion between a pair of intrasubject CT images acquired at different inflation levels and introduce a new vesselness similarity cost that improves intensity-only registration. Volumetric CT datasets from six human subjects were used in this study. The performance of four intensity-only registration algorithms was compared with and without adding the vesselness similarity cost function. Matching accuracy was evaluated using landmarks, vessel tree, and fissure planes. The Jacobian determinant of the transformation was used to reveal the deformation pattern of local parenchymal tissue. The average matching error for intensity-only registration methods was on the order of 1£¿mm at landmarks and 1.5£¿mm on fissure planes. After adding the vesselness preserving cost function, the landmark and fissure positioning errors decreased approximately by 25% and 30%, respectively. The vesselness cost function effectively helped improve the registration accuracy in regions near thoracic cage and near the diaphragm for all the intensity-only registration algorithms tested and also helped produce more consistent and more reliable patterns of regional tissue deformation. 1. Introduction Image registration is used to find the spatial correspondence between two images and plays an important role in pulmonary image analysis. In a sequence of pulmonary scans, image registration provides the spatial locations of corresponding voxels. The computed correspondences describe the motion of the lung between a pair of images at the voxel level. Registration of lung volumes across time or across modalities has been utilized to establish lung atlases [1], estimate regional ventilation and local lung tissue expansion [2¨C5], assess lobar slippage during respiration [6, 7], and measure pulmonary function change following radiation therapy [8]. Lung registration methods are typically intensity-based [2, 5, 9¨C13] or feature-based [14¨C18]. Intensity-based methods consider intensity patterns of the whole lung regions to define similarity measures. They take advantage of the strong contrast between the lung parenchyma and the chest wall, and between the parenchyma, the blood vessels and larger airways. Commonly used intensity-based methods include minimizing intensity difference [3, 10], maximizing mutual information or normalized cross correlation [5, 9], and preserving tissue volume or lung
%U http://www.hindawi.com/journals/ijbi/2012/285136/