Lung Nodule Modeling --- A Data-Driven Approach

Authors:
Amal Farag;James Graham;Aly Farag;Robert Falk
Affiliations:
Department of Electrical and Computer Engineering, University of Louisville,;Department of Electrical and Computer Engineering, University of Louisville,;Department of Electrical and Computer Engineering, University of Louisville,;Department of Electrical and Computer Engineering, University of Louisville,
Venue:
ISVC '09 Proceedings of the 5th International Symposium on Advances in Visual Computing: Part I
Year:
2009

Citing 2
Cited 0

Appearance models for robust segmentation of pulmonary nodules in 3d LDCT chest images

MICCAI'06 Proceedings of the 9th international conference on Medical Image Computing and Computer-Assisted Intervention - Volume Part I
Quantitative nodule detection in low dose chest CT scans: new template modeling and evaluation for CAD system design

MICCAI'05 Proceedings of the 8th international conference on Medical Image Computing and Computer-Assisted Intervention - Volume Part I

Quantified Score

Hi-index	0.00

Visualization

Abstract

The success of lung nodule detection depends on the quality of nodule models. This paper presents a novel approach for nodule modeling which is data-driven. Low dose CT (LDCT) scans of clinical chest data are used to create the required statistics for the models based on modern computer vision techniques. The new models characterize the tissue characteristics of typical lung nodules as specified by human experts. These models suit various machine learning approaches for nodule detection including simulated annealing, genetic algorithms, SVM, AdaBoost, and Bayesian methods. The quality of the new nodule models are studied with respect to parametric models, and are tested on clinical data showing significant improvements in sensitivity and specificity.