Abdominal muscle segmentation from CT using a convolutional neural network

Ka'Toria Edwards; Avneesh Chhabra; James Dormer; Phillip Jones; Robert D. Boutin; Leon Lenchik; Baowei Fei

doi:10.1117/12.2549406

Abdominal muscle segmentation from CT using a convolutional neural network

Ka'Toria Edwards, Avneesh Chhabra, James Dormer, Phillip Jones, Robert D. Boutin, Leon Lenchik, Baowei Fei

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

12 Scopus citations

Abstract

CT is widely used for diagnosis and treatment of a variety of diseases, including characterization of muscle loss. In many cases, changes in muscle mass, particularly abdominal muscle, indicate how well a patient is responding to treatment. Therefore, physicians use CT to monitor changes in muscle mass throughout the patient's course of treatment. In order to measure the muscle, radiologists must segment and review each CT slice manually, which is a time-consuming task. In this work, we present a fully convolutional neural network (CNN) for the segmentation of abdominal muscle on CT. We achieved a mean Dice similarity coefficient of 0.92, a mean precision of 0.93, and a mean recall of 0.91 in an independent test set. The CNN-based segmentation method can provide an automatic tool for the segmentation of abdominal muscle. As a result, the time required to obtain information about changes in abdominal muscle using the CNN takes a fraction of the time associated with manual segmentation methods and thus can provide a useful tool in the clinical application.

Original language	English (US)
Title of host publication	Medical Imaging 2020
Subtitle of host publication	Biomedical Applications in Molecular, Structural, and Functional Imaging
Editors	Andrzej Krol, Barjor S. Gimi
Publisher	SPIE
ISBN (Electronic)	9781510634015
DOIs	https://doi.org/10.1117/12.2549406
State	Published - 2020
Event	Medical Imaging 2020: Biomedical Applications in Molecular, Structural, and Functional Imaging - Houston, United States Duration: Feb 18 2020 → Feb 20 2020

Publication series

Name	Progress in Biomedical Optics and Imaging - Proceedings of SPIE
Volume	11317
ISSN (Print)	1605-7422

Conference

Conference	Medical Imaging 2020: Biomedical Applications in Molecular, Structural, and Functional Imaging
Country/Territory	United States
City	Houston
Period	2/18/20 → 2/20/20

Keywords

CT
Convolutional Neural Networks
Deep Learning
Image segmentation
Muscle Segmentation
Muscle imaging

ASJC Scopus subject areas

Electronic, Optical and Magnetic Materials
Atomic and Molecular Physics, and Optics
Biomaterials
Radiology Nuclear Medicine and imaging

Access to Document

10.1117/12.2549406

Cite this

Edwards, KT., Chhabra, A., Dormer, J., Jones, P., Boutin, R. D., Lenchik, L., & Fei, B. (2020). Abdominal muscle segmentation from CT using a convolutional neural network. In A. Krol, & B. S. Gimi (Eds.), Medical Imaging 2020: Biomedical Applications in Molecular, Structural, and Functional Imaging Article 113170L (Progress in Biomedical Optics and Imaging - Proceedings of SPIE; Vol. 11317). SPIE. https://doi.org/10.1117/12.2549406

Abdominal muscle segmentation from CT using a convolutional neural network. / Edwards, Ka'Toria; Chhabra, Avneesh; Dormer, James et al.
Medical Imaging 2020: Biomedical Applications in Molecular, Structural, and Functional Imaging. ed. / Andrzej Krol; Barjor S. Gimi. SPIE, 2020. 113170L (Progress in Biomedical Optics and Imaging - Proceedings of SPIE; Vol. 11317).

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

Edwards, KT, Chhabra, A, Dormer, J, Jones, P, Boutin, RD, Lenchik, L & Fei, B 2020, Abdominal muscle segmentation from CT using a convolutional neural network. in A Krol & BS Gimi (eds), Medical Imaging 2020: Biomedical Applications in Molecular, Structural, and Functional Imaging., 113170L, Progress in Biomedical Optics and Imaging - Proceedings of SPIE, vol. 11317, SPIE, Medical Imaging 2020: Biomedical Applications in Molecular, Structural, and Functional Imaging, Houston, United States, 2/18/20. https://doi.org/10.1117/12.2549406

Edwards KT, Chhabra A, Dormer J, Jones P, Boutin RD, Lenchik L et al. Abdominal muscle segmentation from CT using a convolutional neural network. In Krol A, Gimi BS, editors, Medical Imaging 2020: Biomedical Applications in Molecular, Structural, and Functional Imaging. SPIE. 2020. 113170L. (Progress in Biomedical Optics and Imaging - Proceedings of SPIE). doi: 10.1117/12.2549406

@inproceedings{4329b53e23ce469ababe32aac86870c4,

title = "Abdominal muscle segmentation from CT using a convolutional neural network",

abstract = "CT is widely used for diagnosis and treatment of a variety of diseases, including characterization of muscle loss. In many cases, changes in muscle mass, particularly abdominal muscle, indicate how well a patient is responding to treatment. Therefore, physicians use CT to monitor changes in muscle mass throughout the patient's course of treatment. In order to measure the muscle, radiologists must segment and review each CT slice manually, which is a time-consuming task. In this work, we present a fully convolutional neural network (CNN) for the segmentation of abdominal muscle on CT. We achieved a mean Dice similarity coefficient of 0.92, a mean precision of 0.93, and a mean recall of 0.91 in an independent test set. The CNN-based segmentation method can provide an automatic tool for the segmentation of abdominal muscle. As a result, the time required to obtain information about changes in abdominal muscle using the CNN takes a fraction of the time associated with manual segmentation methods and thus can provide a useful tool in the clinical application.",

keywords = "CT, Convolutional Neural Networks, Deep Learning, Image segmentation, Muscle Segmentation, Muscle imaging",

author = "Ka'Toria Edwards and Avneesh Chhabra and James Dormer and Phillip Jones and Boutin, {Robert D.} and Leon Lenchik and Baowei Fei",

note = "Funding Information: This research was supported in part by the U.S. National Institutes of Health (NIH) grants (R01CA156775, R01CA204254, R01HL140325, and R21CA231911) and by the Cancer Prevention and Research Institute of Texas (CPRIT) grant RP190588. Publisher Copyright: {\textcopyright} 2020 SPIE; Medical Imaging 2020: Biomedical Applications in Molecular, Structural, and Functional Imaging ; Conference date: 18-02-2020 Through 20-02-2020",

year = "2020",

doi = "10.1117/12.2549406",

language = "English (US)",

series = "Progress in Biomedical Optics and Imaging - Proceedings of SPIE",

publisher = "SPIE",

editor = "Andrzej Krol and Gimi, {Barjor S.}",

booktitle = "Medical Imaging 2020",

}

TY - GEN

T1 - Abdominal muscle segmentation from CT using a convolutional neural network

AU - Edwards, Ka'Toria

AU - Chhabra, Avneesh

AU - Dormer, James

AU - Jones, Phillip

AU - Boutin, Robert D.

AU - Lenchik, Leon

AU - Fei, Baowei

N1 - Funding Information: This research was supported in part by the U.S. National Institutes of Health (NIH) grants (R01CA156775, R01CA204254, R01HL140325, and R21CA231911) and by the Cancer Prevention and Research Institute of Texas (CPRIT) grant RP190588. Publisher Copyright: © 2020 SPIE

PY - 2020

Y1 - 2020

N2 - CT is widely used for diagnosis and treatment of a variety of diseases, including characterization of muscle loss. In many cases, changes in muscle mass, particularly abdominal muscle, indicate how well a patient is responding to treatment. Therefore, physicians use CT to monitor changes in muscle mass throughout the patient's course of treatment. In order to measure the muscle, radiologists must segment and review each CT slice manually, which is a time-consuming task. In this work, we present a fully convolutional neural network (CNN) for the segmentation of abdominal muscle on CT. We achieved a mean Dice similarity coefficient of 0.92, a mean precision of 0.93, and a mean recall of 0.91 in an independent test set. The CNN-based segmentation method can provide an automatic tool for the segmentation of abdominal muscle. As a result, the time required to obtain information about changes in abdominal muscle using the CNN takes a fraction of the time associated with manual segmentation methods and thus can provide a useful tool in the clinical application.

AB - CT is widely used for diagnosis and treatment of a variety of diseases, including characterization of muscle loss. In many cases, changes in muscle mass, particularly abdominal muscle, indicate how well a patient is responding to treatment. Therefore, physicians use CT to monitor changes in muscle mass throughout the patient's course of treatment. In order to measure the muscle, radiologists must segment and review each CT slice manually, which is a time-consuming task. In this work, we present a fully convolutional neural network (CNN) for the segmentation of abdominal muscle on CT. We achieved a mean Dice similarity coefficient of 0.92, a mean precision of 0.93, and a mean recall of 0.91 in an independent test set. The CNN-based segmentation method can provide an automatic tool for the segmentation of abdominal muscle. As a result, the time required to obtain information about changes in abdominal muscle using the CNN takes a fraction of the time associated with manual segmentation methods and thus can provide a useful tool in the clinical application.

KW - CT

KW - Convolutional Neural Networks

KW - Deep Learning

KW - Image segmentation

KW - Muscle Segmentation

KW - Muscle imaging

UR - http://www.scopus.com/inward/record.url?scp=85110492160&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85110492160&partnerID=8YFLogxK

U2 - 10.1117/12.2549406

DO - 10.1117/12.2549406

M3 - Conference contribution

C2 - 32577045

AN - SCOPUS:85110492160

T3 - Progress in Biomedical Optics and Imaging - Proceedings of SPIE

BT - Medical Imaging 2020

A2 - Krol, Andrzej

A2 - Gimi, Barjor S.

PB - SPIE

T2 - Medical Imaging 2020: Biomedical Applications in Molecular, Structural, and Functional Imaging

Y2 - 18 February 2020 through 20 February 2020

ER -

Abdominal muscle segmentation from CT using a convolutional neural network

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this