Accuracy and variability of radiomics in photon-counting CT: Texture features and lung lesion morphology

Jayasai R. Rajagopal; Jocelyn Hoye; Marthony Robins; Elizabeth C. Jones; Ehsan Samei

doi:10.1117/12.2512934

Accuracy and variability of radiomics in photon-counting CT: Texture features and lung lesion morphology

Jayasai R. Rajagopal, Jocelyn Hoye, Marthony Robins, Elizabeth C. Jones, Ehsan Samei

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

1 Scopus citations

Abstract

The purpose of this study was to evaluate the potential of a prototype photon-counting CT system scanner to characterize liver texture and lung lesion morphology features. We utilized a multi-tiered phantom (Mercury Phantom 4.0) to characterize the noise power spectrum and task-transfer functions of both conventional and photoncounting modes on the scanner. Using these metrics, we blurred three textures models and fifteen model lesions for four doses (CTDIvol: 4, 8, 16, 24 mGy), and three slice thicknesses (1.6, 2.5, 4 mm), for a total of 12 imaging conditions. Twenty texture features and twenty-one morphology features were evaluated. Performance was characterized in terms of accuracy (percent bias of features across different conditions) and variability (coefficient of variation of features due to repeats and averaged across conditions). Compared to conventional CT, photon-counting CT had comparable accuracy and variability for texture features. For morphology features, photon-counting CT had comparable accuracy and less variability than conventional CT. For both imaging modes, change in dose showed slight variation in features and increasing slice thickness caused a monotonic change with feature dependent directionality. Photon-counting CT can improve the characterization of morphology features without compromising texture features.

Original language	English
Title of host publication	Medical Imaging 2019
Subtitle of host publication	Physics of Medical Imaging
Editors	Taly Gilat Schmidt, Guang-Hong Chen, Hilde Bosmans
Publisher	SPIE
ISBN (Electronic)	9781510625433
DOIs	https://doi.org/10.1117/12.2512934
State	Published - 2019
Externally published	Yes
Event	Medical Imaging 2019: Physics of Medical Imaging - San Diego, United States Duration: Feb 17 2019 → Feb 20 2019

Publication series

Name	Progress in Biomedical Optics and Imaging - Proceedings of SPIE
Volume	10948
ISSN (Print)	1605-7422

Conference

Conference	Medical Imaging 2019: Physics of Medical Imaging
Country/Territory	United States
City	San Diego
Period	02/17/19 → 02/20/19

Funding

This study was supported by the National Institutes of Health (NIH) Clinical Center Radiology and Imaging Sciences (RADIS), the National Institute of Biomedical Imaging and Bioengineering (NIBIB) and the NIH Intramural Research Program (NIH Z01 1ZID BC011242, and CL040015). The NIH and Siemens Medical Solutions have a Cooperative Research and Development Agreement providing financial and material support including the photon-counting CT system. The content of this manuscript does not necessarily reflect the views or policies of the Department of Health and Human Services, nor do mention of trade names, commercial products, or organizations imply endorsement by the United States Government.

Funders	Funder number
National Institutes of Health
National Institute of Biomedical Imaging and Bioengineering	Z01 1ZID BC011242, CL040015

Keywords

Computed tomography
Morphology
Photon-counting
Quantitative imaging
Texture

Access to Document

10.1117/12.2512934

Cite this

Rajagopal, J. R., Hoye, J., Robins, M., Jones, E. C., & Samei, E. (2019). Accuracy and variability of radiomics in photon-counting CT: Texture features and lung lesion morphology. In T. G. Schmidt, G.-H. Chen, & H. Bosmans (Eds.), Medical Imaging 2019: Physics of Medical Imaging Article 109484J (Progress in Biomedical Optics and Imaging - Proceedings of SPIE; Vol. 10948). SPIE. https://doi.org/10.1117/12.2512934

@inproceedings{7759bd46f9e0487397e83b1f54385716,

title = "Accuracy and variability of radiomics in photon-counting CT: Texture features and lung lesion morphology",

abstract = "The purpose of this study was to evaluate the potential of a prototype photon-counting CT system scanner to characterize liver texture and lung lesion morphology features. We utilized a multi-tiered phantom (Mercury Phantom 4.0) to characterize the noise power spectrum and task-transfer functions of both conventional and photoncounting modes on the scanner. Using these metrics, we blurred three textures models and fifteen model lesions for four doses (CTDIvol: 4, 8, 16, 24 mGy), and three slice thicknesses (1.6, 2.5, 4 mm), for a total of 12 imaging conditions. Twenty texture features and twenty-one morphology features were evaluated. Performance was characterized in terms of accuracy (percent bias of features across different conditions) and variability (coefficient of variation of features due to repeats and averaged across conditions). Compared to conventional CT, photon-counting CT had comparable accuracy and variability for texture features. For morphology features, photon-counting CT had comparable accuracy and less variability than conventional CT. For both imaging modes, change in dose showed slight variation in features and increasing slice thickness caused a monotonic change with feature dependent directionality. Photon-counting CT can improve the characterization of morphology features without compromising texture features.",

keywords = "Computed tomography, Morphology, Photon-counting, Quantitative imaging, Texture",

author = "Rajagopal, {Jayasai R.} and Jocelyn Hoye and Marthony Robins and Jones, {Elizabeth C.} and Ehsan Samei",

note = "Publisher Copyright: {\textcopyright} SPIE. Downloading of the abstract is permitted for personal use only.; Medical Imaging 2019: Physics of Medical Imaging ; Conference date: 17-02-2019 Through 20-02-2019",

year = "2019",

doi = "10.1117/12.2512934",

language = "English",

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

publisher = "SPIE",

editor = "Schmidt, {Taly Gilat} and Guang-Hong Chen and Hilde Bosmans",

booktitle = "Medical Imaging 2019",

}

Rajagopal, JR, Hoye, J, Robins, M, Jones, EC & Samei, E 2019, Accuracy and variability of radiomics in photon-counting CT: Texture features and lung lesion morphology. in TG Schmidt, G-H Chen & H Bosmans (eds), Medical Imaging 2019: Physics of Medical Imaging., 109484J, Progress in Biomedical Optics and Imaging - Proceedings of SPIE, vol. 10948, SPIE, Medical Imaging 2019: Physics of Medical Imaging, San Diego, United States, 02/17/19. https://doi.org/10.1117/12.2512934

Accuracy and variability of radiomics in photon-counting CT: Texture features and lung lesion morphology. / Rajagopal, Jayasai R.; Hoye, Jocelyn; Robins, Marthony et al.
Medical Imaging 2019: Physics of Medical Imaging. ed. / Taly Gilat Schmidt; Guang-Hong Chen; Hilde Bosmans. SPIE, 2019. 109484J (Progress in Biomedical Optics and Imaging - Proceedings of SPIE; Vol. 10948).

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

TY - GEN

T1 - Accuracy and variability of radiomics in photon-counting CT

T2 - Medical Imaging 2019: Physics of Medical Imaging

AU - Rajagopal, Jayasai R.

AU - Hoye, Jocelyn

AU - Robins, Marthony

AU - Jones, Elizabeth C.

AU - Samei, Ehsan

N1 - Publisher Copyright: © SPIE. Downloading of the abstract is permitted for personal use only.

PY - 2019

Y1 - 2019

N2 - The purpose of this study was to evaluate the potential of a prototype photon-counting CT system scanner to characterize liver texture and lung lesion morphology features. We utilized a multi-tiered phantom (Mercury Phantom 4.0) to characterize the noise power spectrum and task-transfer functions of both conventional and photoncounting modes on the scanner. Using these metrics, we blurred three textures models and fifteen model lesions for four doses (CTDIvol: 4, 8, 16, 24 mGy), and three slice thicknesses (1.6, 2.5, 4 mm), for a total of 12 imaging conditions. Twenty texture features and twenty-one morphology features were evaluated. Performance was characterized in terms of accuracy (percent bias of features across different conditions) and variability (coefficient of variation of features due to repeats and averaged across conditions). Compared to conventional CT, photon-counting CT had comparable accuracy and variability for texture features. For morphology features, photon-counting CT had comparable accuracy and less variability than conventional CT. For both imaging modes, change in dose showed slight variation in features and increasing slice thickness caused a monotonic change with feature dependent directionality. Photon-counting CT can improve the characterization of morphology features without compromising texture features.

AB - The purpose of this study was to evaluate the potential of a prototype photon-counting CT system scanner to characterize liver texture and lung lesion morphology features. We utilized a multi-tiered phantom (Mercury Phantom 4.0) to characterize the noise power spectrum and task-transfer functions of both conventional and photoncounting modes on the scanner. Using these metrics, we blurred three textures models and fifteen model lesions for four doses (CTDIvol: 4, 8, 16, 24 mGy), and three slice thicknesses (1.6, 2.5, 4 mm), for a total of 12 imaging conditions. Twenty texture features and twenty-one morphology features were evaluated. Performance was characterized in terms of accuracy (percent bias of features across different conditions) and variability (coefficient of variation of features due to repeats and averaged across conditions). Compared to conventional CT, photon-counting CT had comparable accuracy and variability for texture features. For morphology features, photon-counting CT had comparable accuracy and less variability than conventional CT. For both imaging modes, change in dose showed slight variation in features and increasing slice thickness caused a monotonic change with feature dependent directionality. Photon-counting CT can improve the characterization of morphology features without compromising texture features.

KW - Computed tomography

KW - Morphology

KW - Photon-counting

KW - Quantitative imaging

KW - Texture

UR - http://www.scopus.com/inward/record.url?scp=85068380378&partnerID=8YFLogxK

U2 - 10.1117/12.2512934

DO - 10.1117/12.2512934

M3 - Conference contribution

AN - SCOPUS:85068380378

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

BT - Medical Imaging 2019

A2 - Schmidt, Taly Gilat

A2 - Chen, Guang-Hong

A2 - Bosmans, Hilde

PB - SPIE

Y2 - 17 February 2019 through 20 February 2019

ER -

Accuracy and variability of radiomics in photon-counting CT: Texture features and lung lesion morphology

Abstract

Publication series

Conference

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this