Characterization of lung nodule malignancy using hybrid shape and appearance features

Mario Buty; Ziyue Xu; Mingchen Gao; Ulas Bagci; Aaron Wu; Daniel J. Mollura

doi:10.1007/978-3-319-46720-7_77

Characterization of lung nodule malignancy using hybrid shape and appearance features

Mario Buty
, Ziyue Xu
, Mingchen Gao
, Ulas Bagci
, Aaron Wu
, Daniel J. Mollura

Department of Computer Science and Engineering

National Institutes of Health
University of Central Florida

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

50 Scopus citations

Abstract

Computed tomography imaging is a standard modality for detecting and assessing lung cancer. In order to evaluate the malignancy of lung nodules,clinical practice often involves expert qualitative ratings on several criteria describing a nodule’s appearance and shape. Translating these features for computer-aided diagnostics is challenging due to their subjective nature and the difficulties in gaining a complete description. In this paper,we propose a computerized approach to quantitatively evaluate both appearance distinctions and 3D surface variations. Nodule shape was modeled and parameterized using spherical harmonics,and appearance features were extracted using deep convolutional neural networks. Both sets of features were combined to estimate the nodule malignancy using a random forest classifier. The proposed algorithm was tested on the publicly available Lung Image Database Consortium dataset,achieving high accuracy. By providing lung nodule characterization,this method can provide a robust alternative reference opinion for lung cancer diagnosis.

Original language	English
Title of host publication	Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings
Editors	Sebastian Ourselin, Leo Joskowicz, Mert R. Sabuncu, William Wells, Gozde Unal
Publisher	Springer Verlag
Pages	662-670
Number of pages	9
ISBN (Print)	9783319467191
DOIs	https://doi.org/10.1007/978-3-319-46720-7_77
State	Published - 2016
Event	1st International Workshop on Simulation and Synthesis in Medical Imaging, SASHIMI 2016 held in conjunction with 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016 - Athens, Greece Duration: Oct 21 2016 → Oct 21 2016

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	9900 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	1st International Workshop on Simulation and Synthesis in Medical Imaging, SASHIMI 2016 held in conjunction with 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016
Country/Territory	Greece
City	Athens
Period	10/21/16 → 10/21/16

Keywords

Conformal mapping
Deep convolutional neural network
Nodule characterization
Random forest
Spherical harmonics

Access to Document

10.1007/978-3-319-46720-7_77

Cite this

Buty, M., Xu, Z., Gao, M., Bagci, U., Wu, A., & Mollura, D. J. (2016). Characterization of lung nodule malignancy using hybrid shape and appearance features. In S. Ourselin, L. Joskowicz, M. R. Sabuncu, W. Wells, & G. Unal (Eds.), Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings (pp. 662-670). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9900 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-46720-7_77

Buty, Mario ; Xu, Ziyue ; Gao, Mingchen et al. / Characterization of lung nodule malignancy using hybrid shape and appearance features. Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings. editor / Sebastian Ourselin ; Leo Joskowicz ; Mert R. Sabuncu ; William Wells ; Gozde Unal. Springer Verlag, 2016. pp. 662-670 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{499103a25af54f54a6f2c0cfcdcd0263,

title = "Characterization of lung nodule malignancy using hybrid shape and appearance features",

abstract = "Computed tomography imaging is a standard modality for detecting and assessing lung cancer. In order to evaluate the malignancy of lung nodules,clinical practice often involves expert qualitative ratings on several criteria describing a nodule{\textquoteright}s appearance and shape. Translating these features for computer-aided diagnostics is challenging due to their subjective nature and the difficulties in gaining a complete description. In this paper,we propose a computerized approach to quantitatively evaluate both appearance distinctions and 3D surface variations. Nodule shape was modeled and parameterized using spherical harmonics,and appearance features were extracted using deep convolutional neural networks. Both sets of features were combined to estimate the nodule malignancy using a random forest classifier. The proposed algorithm was tested on the publicly available Lung Image Database Consortium dataset,achieving high accuracy. By providing lung nodule characterization,this method can provide a robust alternative reference opinion for lung cancer diagnosis.",

keywords = "Conformal mapping, Deep convolutional neural network, Nodule characterization, Random forest, Spherical harmonics",

author = "Mario Buty and Ziyue Xu and Mingchen Gao and Ulas Bagci and Aaron Wu and Mollura, \{Daniel J.\}",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing AG 2016.; 1st International Workshop on Simulation and Synthesis in Medical Imaging, SASHIMI 2016 held in conjunction with 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016 ; Conference date: 21-10-2016 Through 21-10-2016",

year = "2016",

doi = "10.1007/978-3-319-46720-7\_77",

language = "English",

isbn = "9783319467191",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "662--670",

editor = "Sebastian Ourselin and Leo Joskowicz and Sabuncu, \{Mert R.\} and William Wells and Gozde Unal",

booktitle = "Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings",

address = "Germany",

}

Buty, M, Xu, Z, Gao, M, Bagci, U, Wu, A & Mollura, DJ 2016, Characterization of lung nodule malignancy using hybrid shape and appearance features. in S Ourselin, L Joskowicz, MR Sabuncu, W Wells & G Unal (eds), Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9900 LNCS, Springer Verlag, pp. 662-670, 1st International Workshop on Simulation and Synthesis in Medical Imaging, SASHIMI 2016 held in conjunction with 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016, Athens, Greece, 10/21/16. https://doi.org/10.1007/978-3-319-46720-7_77

Characterization of lung nodule malignancy using hybrid shape and appearance features. / Buty, Mario; Xu, Ziyue; Gao, Mingchen et al.
Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings. ed. / Sebastian Ourselin; Leo Joskowicz; Mert R. Sabuncu; William Wells; Gozde Unal. Springer Verlag, 2016. p. 662-670 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9900 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Characterization of lung nodule malignancy using hybrid shape and appearance features

AU - Buty, Mario

AU - Xu, Ziyue

AU - Gao, Mingchen

AU - Bagci, Ulas

AU - Wu, Aaron

AU - Mollura, Daniel J.

N1 - Publisher Copyright: © Springer International Publishing AG 2016.

PY - 2016

Y1 - 2016

N2 - Computed tomography imaging is a standard modality for detecting and assessing lung cancer. In order to evaluate the malignancy of lung nodules,clinical practice often involves expert qualitative ratings on several criteria describing a nodule’s appearance and shape. Translating these features for computer-aided diagnostics is challenging due to their subjective nature and the difficulties in gaining a complete description. In this paper,we propose a computerized approach to quantitatively evaluate both appearance distinctions and 3D surface variations. Nodule shape was modeled and parameterized using spherical harmonics,and appearance features were extracted using deep convolutional neural networks. Both sets of features were combined to estimate the nodule malignancy using a random forest classifier. The proposed algorithm was tested on the publicly available Lung Image Database Consortium dataset,achieving high accuracy. By providing lung nodule characterization,this method can provide a robust alternative reference opinion for lung cancer diagnosis.

AB - Computed tomography imaging is a standard modality for detecting and assessing lung cancer. In order to evaluate the malignancy of lung nodules,clinical practice often involves expert qualitative ratings on several criteria describing a nodule’s appearance and shape. Translating these features for computer-aided diagnostics is challenging due to their subjective nature and the difficulties in gaining a complete description. In this paper,we propose a computerized approach to quantitatively evaluate both appearance distinctions and 3D surface variations. Nodule shape was modeled and parameterized using spherical harmonics,and appearance features were extracted using deep convolutional neural networks. Both sets of features were combined to estimate the nodule malignancy using a random forest classifier. The proposed algorithm was tested on the publicly available Lung Image Database Consortium dataset,achieving high accuracy. By providing lung nodule characterization,this method can provide a robust alternative reference opinion for lung cancer diagnosis.

KW - Conformal mapping

KW - Deep convolutional neural network

KW - Nodule characterization

KW - Random forest

KW - Spherical harmonics

UR - https://www.scopus.com/pages/publications/84996503636

U2 - 10.1007/978-3-319-46720-7_77

DO - 10.1007/978-3-319-46720-7_77

M3 - Conference contribution

AN - SCOPUS:84996503636

SN - 9783319467191

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 662

EP - 670

BT - Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings

A2 - Ourselin, Sebastian

A2 - Joskowicz, Leo

A2 - Sabuncu, Mert R.

A2 - Wells, William

A2 - Unal, Gozde

PB - Springer Verlag

T2 - 1st International Workshop on Simulation and Synthesis in Medical Imaging, SASHIMI 2016 held in conjunction with 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016

Y2 - 21 October 2016 through 21 October 2016

ER -

Buty M, Xu Z, Gao M, Bagci U, Wu A, Mollura DJ. Characterization of lung nodule malignancy using hybrid shape and appearance features. In Ourselin S, Joskowicz L, Sabuncu MR, Wells W, Unal G, editors, Medical Image Computing and Computer-Assisted Intervention - MICCAI 2016 - 19th International Conference, Proceedings. Springer Verlag. 2016. p. 662-670. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-46720-7_77

Characterization of lung nodule malignancy using hybrid shape and appearance features

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this