Lung nodule classification with multilevel patch-based context analysis

Fan Zhang, Yang Song, Weidong Cai, Min Zhao Lee, Yun Zhou, Heng Huang, Shimin Shan, Michael J. Fulham, Dagan D. Feng

Research output: Contribution to journalArticle

Abstract

In this paper, we propose a novel classification method for the four types of lung nodules, i.e., well-circumscribed, vascularized, juxta-pleural, and pleural-tail, in low dose computed tomography scans. The proposed method is based on contextual analysis by combining the lung nodule and surrounding anatomical structures, and has three main stages: an adaptive patch-based division is used to construct concentric multilevel partition; then, a new feature set is designed to incorporate intensity, texture, and gradient information for image patch feature description, and then a contextual latent semantic analysis-based classifier is designed to calculate the probabilistic estimations for the relevant images. Our proposed method was evaluated on a publicly available dataset and clearly demonstrated promising classification performance.

Original languageEnglish (US)
Article number6690248
Pages (from-to)1155-1166
Number of pages12
JournalIEEE Transactions on Biomedical Engineering
Volume61
Issue number4
DOIs
StatePublished - 2014

Fingerprint

Tomography
Classifiers
Textures
Semantics

Keywords

  • Classification
  • feature design
  • latent semantic analysis
  • patch division

ASJC Scopus subject areas

  • Biomedical Engineering

Cite this

Zhang, F., Song, Y., Cai, W., Lee, M. Z., Zhou, Y., Huang, H., ... Feng, D. D. (2014). Lung nodule classification with multilevel patch-based context analysis. IEEE Transactions on Biomedical Engineering, 61(4), 1155-1166. [6690248]. https://doi.org/10.1109/TBME.2013.2295593

Lung nodule classification with multilevel patch-based context analysis. / Zhang, Fan; Song, Yang; Cai, Weidong; Lee, Min Zhao; Zhou, Yun; Huang, Heng; Shan, Shimin; Fulham, Michael J.; Feng, Dagan D.

In: IEEE Transactions on Biomedical Engineering, Vol. 61, No. 4, 6690248, 2014, p. 1155-1166.

Research output: Contribution to journalArticle

Zhang, F, Song, Y, Cai, W, Lee, MZ, Zhou, Y, Huang, H, Shan, S, Fulham, MJ & Feng, DD 2014, 'Lung nodule classification with multilevel patch-based context analysis', IEEE Transactions on Biomedical Engineering, vol. 61, no. 4, 6690248, pp. 1155-1166. https://doi.org/10.1109/TBME.2013.2295593
Zhang, Fan ; Song, Yang ; Cai, Weidong ; Lee, Min Zhao ; Zhou, Yun ; Huang, Heng ; Shan, Shimin ; Fulham, Michael J. ; Feng, Dagan D. / Lung nodule classification with multilevel patch-based context analysis. In: IEEE Transactions on Biomedical Engineering. 2014 ; Vol. 61, No. 4. pp. 1155-1166.
@article{29e9d7f3e71f46aa834e04c14575e398,
title = "Lung nodule classification with multilevel patch-based context analysis",
abstract = "In this paper, we propose a novel classification method for the four types of lung nodules, i.e., well-circumscribed, vascularized, juxta-pleural, and pleural-tail, in low dose computed tomography scans. The proposed method is based on contextual analysis by combining the lung nodule and surrounding anatomical structures, and has three main stages: an adaptive patch-based division is used to construct concentric multilevel partition; then, a new feature set is designed to incorporate intensity, texture, and gradient information for image patch feature description, and then a contextual latent semantic analysis-based classifier is designed to calculate the probabilistic estimations for the relevant images. Our proposed method was evaluated on a publicly available dataset and clearly demonstrated promising classification performance.",
keywords = "Classification, feature design, latent semantic analysis, patch division",
author = "Fan Zhang and Yang Song and Weidong Cai and Lee, {Min Zhao} and Yun Zhou and Heng Huang and Shimin Shan and Fulham, {Michael J.} and Feng, {Dagan D.}",
year = "2014",
doi = "10.1109/TBME.2013.2295593",
language = "English (US)",
volume = "61",
pages = "1155--1166",
journal = "IEEE Transactions on Biomedical Engineering",
issn = "0018-9294",
publisher = "IEEE Computer Society",
number = "4",

}

TY - JOUR

T1 - Lung nodule classification with multilevel patch-based context analysis

AU - Zhang, Fan

AU - Song, Yang

AU - Cai, Weidong

AU - Lee, Min Zhao

AU - Zhou, Yun

AU - Huang, Heng

AU - Shan, Shimin

AU - Fulham, Michael J.

AU - Feng, Dagan D.

PY - 2014

Y1 - 2014

N2 - In this paper, we propose a novel classification method for the four types of lung nodules, i.e., well-circumscribed, vascularized, juxta-pleural, and pleural-tail, in low dose computed tomography scans. The proposed method is based on contextual analysis by combining the lung nodule and surrounding anatomical structures, and has three main stages: an adaptive patch-based division is used to construct concentric multilevel partition; then, a new feature set is designed to incorporate intensity, texture, and gradient information for image patch feature description, and then a contextual latent semantic analysis-based classifier is designed to calculate the probabilistic estimations for the relevant images. Our proposed method was evaluated on a publicly available dataset and clearly demonstrated promising classification performance.

AB - In this paper, we propose a novel classification method for the four types of lung nodules, i.e., well-circumscribed, vascularized, juxta-pleural, and pleural-tail, in low dose computed tomography scans. The proposed method is based on contextual analysis by combining the lung nodule and surrounding anatomical structures, and has three main stages: an adaptive patch-based division is used to construct concentric multilevel partition; then, a new feature set is designed to incorporate intensity, texture, and gradient information for image patch feature description, and then a contextual latent semantic analysis-based classifier is designed to calculate the probabilistic estimations for the relevant images. Our proposed method was evaluated on a publicly available dataset and clearly demonstrated promising classification performance.

KW - Classification

KW - feature design

KW - latent semantic analysis

KW - patch division

UR - http://www.scopus.com/inward/record.url?scp=84897470683&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84897470683&partnerID=8YFLogxK

U2 - 10.1109/TBME.2013.2295593

DO - 10.1109/TBME.2013.2295593

M3 - Article

C2 - 24658240

AN - SCOPUS:84897470683

VL - 61

SP - 1155

EP - 1166

JO - IEEE Transactions on Biomedical Engineering

JF - IEEE Transactions on Biomedical Engineering

SN - 0018-9294

IS - 4

M1 - 6690248

ER -