Automated segmentation framework of lung gross tumor volumes on 3D planning CT images using dense V-Net deep learning

Risa Nakano; Hidetaka Arimura; Mohammad Haekal; Saiji Ohga

doi:10.1117/12.2521509

Automated segmentation framework of lung gross tumor volumes on 3D planning CT images using dense V-Net deep learning

Risa Nakano^*
, Hidetaka Arimura
, Mohammad Haekal
, Saiji Ohga

^*Corresponding author for this work

Kyushu University

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

1 Citation (Scopus)

Abstract

Gross tumor volume (GTV) regions of lung tumors should be determined with repeatability and reproducibility on planning computed tomography (CT) in radiation treatment planning to reduce intra- and inter-observer variations of GTV regions. Therefore, we have attempted to develop an automated segmentation framework of the GTV regions on planning CT images using dense V-Net deep learning (DenseVDL). In order to evaluate the GTV regions extracted by the DenseVDL network, Dice similarity coefficient (DSC) was used in this study. The proposed framework achieved average 2D-DSC of 0.73 and 3D-DSC of 0.76 for sixteen cases. The proposed framework using the DenseVDL may be useful for assisting in radiation treatment planning for lung cancer.

Original language	English
Title of host publication	International Forum on Medical Imaging in Asia 2019
Editors	Jong Hyo Kim, Feng Lin, Hiroshi Fujita
Publisher	SPIE
ISBN (Electronic)	9781510627758
DOIs	https://doi.org/10.1117/12.2521509
Publication status	Published - 2019
Externally published	Yes
Event	International Forum on Medical Imaging in Asia 2019 - Singapore, Singapore Duration: 7 Jan 2019 → 9 Jan 2019

Publication series

Name	Proceedings of SPIE - The International Society for Optical Engineering
Volume	11050
ISSN (Print)	0277-786X
ISSN (Electronic)	1996-756X

Conference

Conference	International Forum on Medical Imaging in Asia 2019
Country/Territory	Singapore
City	Singapore
Period	7/01/19 → 9/01/19

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being

Keywords

3D-medical image
Deep learning
Segmentation
dense V-Net

Access to Document

10.1117/12.2521509

Cite this

Nakano, R., Arimura, H., Haekal, M., & Ohga, S. (2019). Automated segmentation framework of lung gross tumor volumes on 3D planning CT images using dense V-Net deep learning. In J. H. Kim, F. Lin, & H. Fujita (Eds.), International Forum on Medical Imaging in Asia 2019 Article 110500Y (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 11050). SPIE. https://doi.org/10.1117/12.2521509

Nakano, Risa ; Arimura, Hidetaka ; Haekal, Mohammad et al. / Automated segmentation framework of lung gross tumor volumes on 3D planning CT images using dense V-Net deep learning. International Forum on Medical Imaging in Asia 2019. editor / Jong Hyo Kim ; Feng Lin ; Hiroshi Fujita. SPIE, 2019. (Proceedings of SPIE - The International Society for Optical Engineering).

@inproceedings{48bcb395982c461894b11100ce42b5e3,

title = "Automated segmentation framework of lung gross tumor volumes on 3D planning CT images using dense V-Net deep learning",

abstract = "Gross tumor volume (GTV) regions of lung tumors should be determined with repeatability and reproducibility on planning computed tomography (CT) in radiation treatment planning to reduce intra- and inter-observer variations of GTV regions. Therefore, we have attempted to develop an automated segmentation framework of the GTV regions on planning CT images using dense V-Net deep learning (DenseVDL). In order to evaluate the GTV regions extracted by the DenseVDL network, Dice similarity coefficient (DSC) was used in this study. The proposed framework achieved average 2D-DSC of 0.73 and 3D-DSC of 0.76 for sixteen cases. The proposed framework using the DenseVDL may be useful for assisting in radiation treatment planning for lung cancer.",

keywords = "3D-medical image, Deep learning, Segmentation, dense V-Net",

author = "Risa Nakano and Hidetaka Arimura and Mohammad Haekal and Saiji Ohga",

note = "Publisher Copyright: {\textcopyright} 2019 SPIE.; International Forum on Medical Imaging in Asia 2019 ; Conference date: 07-01-2019 Through 09-01-2019",

year = "2019",

doi = "10.1117/12.2521509",

language = "English",

series = "Proceedings of SPIE - The International Society for Optical Engineering",

publisher = "SPIE",

editor = "Kim, \{Jong Hyo\} and Feng Lin and Hiroshi Fujita",

booktitle = "International Forum on Medical Imaging in Asia 2019",

address = "United States",

}

Nakano, R, Arimura, H, Haekal, M & Ohga, S 2019, Automated segmentation framework of lung gross tumor volumes on 3D planning CT images using dense V-Net deep learning. in JH Kim, F Lin & H Fujita (eds), International Forum on Medical Imaging in Asia 2019., 110500Y, Proceedings of SPIE - The International Society for Optical Engineering, vol. 11050, SPIE, International Forum on Medical Imaging in Asia 2019, Singapore, Singapore, 7/01/19. https://doi.org/10.1117/12.2521509

Automated segmentation framework of lung gross tumor volumes on 3D planning CT images using dense V-Net deep learning. / Nakano, Risa; Arimura, Hidetaka; Haekal, Mohammad et al.
International Forum on Medical Imaging in Asia 2019. ed. / Jong Hyo Kim; Feng Lin; Hiroshi Fujita. SPIE, 2019. 110500Y (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 11050).

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

TY - GEN

T1 - Automated segmentation framework of lung gross tumor volumes on 3D planning CT images using dense V-Net deep learning

AU - Nakano, Risa

AU - Arimura, Hidetaka

AU - Haekal, Mohammad

AU - Ohga, Saiji

PY - 2019

Y1 - 2019

N2 - Gross tumor volume (GTV) regions of lung tumors should be determined with repeatability and reproducibility on planning computed tomography (CT) in radiation treatment planning to reduce intra- and inter-observer variations of GTV regions. Therefore, we have attempted to develop an automated segmentation framework of the GTV regions on planning CT images using dense V-Net deep learning (DenseVDL). In order to evaluate the GTV regions extracted by the DenseVDL network, Dice similarity coefficient (DSC) was used in this study. The proposed framework achieved average 2D-DSC of 0.73 and 3D-DSC of 0.76 for sixteen cases. The proposed framework using the DenseVDL may be useful for assisting in radiation treatment planning for lung cancer.

AB - Gross tumor volume (GTV) regions of lung tumors should be determined with repeatability and reproducibility on planning computed tomography (CT) in radiation treatment planning to reduce intra- and inter-observer variations of GTV regions. Therefore, we have attempted to develop an automated segmentation framework of the GTV regions on planning CT images using dense V-Net deep learning (DenseVDL). In order to evaluate the GTV regions extracted by the DenseVDL network, Dice similarity coefficient (DSC) was used in this study. The proposed framework achieved average 2D-DSC of 0.73 and 3D-DSC of 0.76 for sixteen cases. The proposed framework using the DenseVDL may be useful for assisting in radiation treatment planning for lung cancer.

KW - 3D-medical image

KW - Deep learning

KW - Segmentation

KW - dense V-Net

UR - https://www.scopus.com/pages/publications/85063903370

U2 - 10.1117/12.2521509

DO - 10.1117/12.2521509

M3 - Conference contribution

AN - SCOPUS:85063903370

T3 - Proceedings of SPIE - The International Society for Optical Engineering

BT - International Forum on Medical Imaging in Asia 2019

A2 - Kim, Jong Hyo

A2 - Lin, Feng

A2 - Fujita, Hiroshi

PB - SPIE

T2 - International Forum on Medical Imaging in Asia 2019

Y2 - 7 January 2019 through 9 January 2019

ER -

Automated segmentation framework of lung gross tumor volumes on 3D planning CT images using dense V-Net deep learning

Abstract

Publication series

Conference

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this