Abstract | ||
---|---|---|
The automatic detection of lung nodules attached to other pulmonary structures is a useful yet challenging task in lung CAD systems. In this paper, we propose a stratified statistical learning approach to recognize whether a candidate nodule detected in CT images connects to any of three other major lung anatomies, namely vessel, fissure and lung wall, or is solitary with background parenchyma. First, we develop a fully automated voxel-by-voxel labeling/segmentation method of nodule, vessel, fissure, lung wall and parenchyma given a 3D lung image, via a unified feature set and classifier under conditional random field. Second, the generated Class Probability Response Maps (PRM) by voxel-level classifiers, are used to form the so-called pairwise Probability Co-occurrence Maps (PCM) which encode the spatial contextual correlations of the candidate nodule, in relation to other anatomical landmarks. Based on PCMs, higher level classifiers are trained to recognize whether the nodule touches other pulmonary structures, as a multi-label problem. We also present a new iterative fissure structure enhancement filter with superior performance. For experimental validation, we create an annotated database of 784 subvolumes with nodules of various sizes, shapes, densities and contextual anatomies, and from 239 patients. High accuracy of multi-class voxel labeling is achieved 89.3% ~ 91.2%. The Area under ROC Curve (AUC) of vessel, fissure and lung wall connectivity classification reaches 0.8676, 0.8692 and 0.9275, respectively. |
Year | DOI | Venue |
---|---|---|
2010 | 10.1109/CVPR.2010.5540008 | CVPR |
Keywords | Field | DocType |
computerised tomography,random processes,automated voxel-by-voxel labeling-segmentation method,statistical analysis,pairwise probability co-occurrence maps,learning (artificial intelligence),image segmentation,stratified statistical learning,curve fitting,class probability response map,ct images,image classification,multiclass voxel labeling,conditional random field,3d lung image,lung nodules automatic detection,spatial contextual correlations,voxel-level classifiers,filtering theory,local anatomical context,area under roc curve,iterative fissure structure enhancement filter,image enhancement,background parenchyma,cad systems,pulmonary structures,correlation methods,iterative methods,medical image processing,shape,computed tomography,cancer,databases,learning artificial intelligence,image recognition,anatomy,labeling,feature extraction | Conditional random field,Voxel,Computer vision,Pattern recognition,Segmentation,Computer science,Feature extraction,Image segmentation,Artificial intelligence,Classifier (linguistics),Fissure,Contextual image classification | Conference |
Volume | Issue | ISSN |
2010 | 1 | 1063-6919 |
ISBN | Citations | PageRank |
978-1-4244-6984-0 | 21 | 0.88 |
References | Authors | |
22 | 7 |
Name | Order | Citations | PageRank |
---|---|---|---|
Dijia Wu | 1 | 102 | 13.75 |
Le Lu | 2 | 1297 | 86.78 |
Jinbo Bi | 3 | 1432 | 104.24 |
Yoshihisa Shinagawa | 4 | 1900 | 124.80 |
Kim L. Boyer | 5 | 1514 | 254.20 |
Arun Krishnan | 6 | 220 | 14.86 |
Marcos Salganicoff | 7 | 307 | 32.32 |