Noise reduction method in low-dose CT data combining neural networks and an iterative reconstruction technique

Dahim Choi; Juhee Kim; Seung Hoon Chae; Jongduk Baek; Andreas Maier; Rebecca Fahrig; Hyun Seok Park; Jang Hwan Choi

doi:10.1117/12.2521445

Noise reduction method in low-dose CT data combining neural networks and an iterative reconstruction technique

Dahim Choi, Juhee Kim, Seung Hoon Chae, Jongduk Baek, Andreas Maier, Rebecca Fahrig, Hyun Seok Park, Jang Hwan Choi

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

1 Scopus citations

Abstract

Improving image quality from low-dose CT image and keeping diagnostic features is integral to lowering the amount of exposure to radiation and its potential risks. Noise reduction methods using deep neural network have been developed and displayed impressive performance, but there are limitations on noise remnants, blurring on high-frequency edge, and artifacts occurrence. To increase noise reduction performance and deal with those issues simultaneously, we have implemented block-based REDCNN model and applied patch-based Landweber-type iteration to images passed through REDCNN model. The model successfully smooths noise on CT images which are imposed Gaussian and Poisson noise, and outperforms noise reduction by other state-of-the-art deep neural network models. We also have tested the effect of repetition of an iterative reconstruction, changing a step size and the number of iteration.

Original language	English
Title of host publication	International Forum on Medical Imaging in Asia 2019
Editors	Feng Lin, Hiroshi Fujita, Jong Hyo Kim
Publisher	SPIE
ISBN (Electronic)	9781510627758
DOIs	https://doi.org/10.1117/12.2521445
State	Published - 2019
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

Bibliographical note

Funding Information:
This research was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government, MSIP (grant no: NRF-2017R1C1B5018287, NRF-2015M3A9A7029725, and NRF-2017M2A2A6A02070522, URL: http://nrf.re.kr). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Publisher Copyright:
© 2019 SPIE.

Keywords

Deep Neural Networks
Landweber iteration
Low-dose CT
Noise Reduction
REDCNN

Access to Document

10.1117/12.2521445

Cite this

Choi, D., Kim, J., Chae, S. H., Baek, J., Maier, A., Fahrig, R., Park, H. S., & Choi, J. H. (2019). Noise reduction method in low-dose CT data combining neural networks and an iterative reconstruction technique. In F. Lin, H. Fujita, & J. H. Kim (Eds.), International Forum on Medical Imaging in Asia 2019 Article 110500C (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 11050). SPIE. https://doi.org/10.1117/12.2521445

@inproceedings{eab1f6087c574b67bacbceda77fdf3f4,

title = "Noise reduction method in low-dose CT data combining neural networks and an iterative reconstruction technique",

abstract = "Improving image quality from low-dose CT image and keeping diagnostic features is integral to lowering the amount of exposure to radiation and its potential risks. Noise reduction methods using deep neural network have been developed and displayed impressive performance, but there are limitations on noise remnants, blurring on high-frequency edge, and artifacts occurrence. To increase noise reduction performance and deal with those issues simultaneously, we have implemented block-based REDCNN model and applied patch-based Landweber-type iteration to images passed through REDCNN model. The model successfully smooths noise on CT images which are imposed Gaussian and Poisson noise, and outperforms noise reduction by other state-of-the-art deep neural network models. We also have tested the effect of repetition of an iterative reconstruction, changing a step size and the number of iteration.",

keywords = "Deep Neural Networks, Landweber iteration, Low-dose CT, Noise Reduction, REDCNN",

author = "Dahim Choi and Juhee Kim and Chae, {Seung Hoon} and Jongduk Baek and Andreas Maier and Rebecca Fahrig and Park, {Hyun Seok} and Choi, {Jang Hwan}",

note = "Funding Information: This research was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government, MSIP (grant no: NRF-2017R1C1B5018287, NRF-2015M3A9A7029725, and NRF-2017M2A2A6A02070522, URL: http://nrf.re.kr). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. 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 = "10.1117/12.2521445",

language = "English",

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

publisher = "SPIE",

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

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

}

Choi, D, Kim, J, Chae, SH, Baek, J, Maier, A, Fahrig, R, Park, HS & Choi, JH 2019, Noise reduction method in low-dose CT data combining neural networks and an iterative reconstruction technique. in F Lin, H Fujita & JH Kim (eds), International Forum on Medical Imaging in Asia 2019., 110500C, 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/10.1117/12.2521445

Noise reduction method in low-dose CT data combining neural networks and an iterative reconstruction technique. / Choi, Dahim; Kim, Juhee; Chae, Seung Hoon et al.
International Forum on Medical Imaging in Asia 2019. ed. / Feng Lin; Hiroshi Fujita; Jong Hyo Kim. SPIE, 2019. 110500C (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 - Noise reduction method in low-dose CT data combining neural networks and an iterative reconstruction technique

AU - Choi, Dahim

AU - Kim, Juhee

AU - Chae, Seung Hoon

AU - Baek, Jongduk

AU - Maier, Andreas

AU - Fahrig, Rebecca

AU - Park, Hyun Seok

AU - Choi, Jang Hwan

N1 - Funding Information: This research was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government, MSIP (grant no: NRF-2017R1C1B5018287, NRF-2015M3A9A7029725, and NRF-2017M2A2A6A02070522, URL: http://nrf.re.kr). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Publisher Copyright: © 2019 SPIE.

PY - 2019

Y1 - 2019

N2 - Improving image quality from low-dose CT image and keeping diagnostic features is integral to lowering the amount of exposure to radiation and its potential risks. Noise reduction methods using deep neural network have been developed and displayed impressive performance, but there are limitations on noise remnants, blurring on high-frequency edge, and artifacts occurrence. To increase noise reduction performance and deal with those issues simultaneously, we have implemented block-based REDCNN model and applied patch-based Landweber-type iteration to images passed through REDCNN model. The model successfully smooths noise on CT images which are imposed Gaussian and Poisson noise, and outperforms noise reduction by other state-of-the-art deep neural network models. We also have tested the effect of repetition of an iterative reconstruction, changing a step size and the number of iteration.

AB - Improving image quality from low-dose CT image and keeping diagnostic features is integral to lowering the amount of exposure to radiation and its potential risks. Noise reduction methods using deep neural network have been developed and displayed impressive performance, but there are limitations on noise remnants, blurring on high-frequency edge, and artifacts occurrence. To increase noise reduction performance and deal with those issues simultaneously, we have implemented block-based REDCNN model and applied patch-based Landweber-type iteration to images passed through REDCNN model. The model successfully smooths noise on CT images which are imposed Gaussian and Poisson noise, and outperforms noise reduction by other state-of-the-art deep neural network models. We also have tested the effect of repetition of an iterative reconstruction, changing a step size and the number of iteration.

KW - Deep Neural Networks

KW - Landweber iteration

KW - Low-dose CT

KW - Noise Reduction

KW - REDCNN

UR - http://www.scopus.com/inward/record.url?scp=85063879382&partnerID=8YFLogxK

U2 - 10.1117/12.2521445

DO - 10.1117/12.2521445

M3 - Conference contribution

AN - SCOPUS:85063879382

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

BT - International Forum on Medical Imaging in Asia 2019

A2 - Lin, Feng

A2 - Fujita, Hiroshi

A2 - Kim, Jong Hyo

PB - SPIE

T2 - International Forum on Medical Imaging in Asia 2019

Y2 - 7 January 2019 through 9 January 2019

ER -

Choi D, Kim J, Chae SH, Baek J, Maier A, Fahrig R et al. Noise reduction method in low-dose CT data combining neural networks and an iterative reconstruction technique. In Lin F, Fujita H, Kim JH, editors, International Forum on Medical Imaging in Asia 2019. SPIE. 2019. 110500C. (Proceedings of SPIE - The International Society for Optical Engineering). doi: 10.1117/12.2521445

Noise reduction method in low-dose CT data combining neural networks and an iterative reconstruction technique

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Fingerprint

Cite this