Denoising X-ray CT images based on product Gaussian mixture distribution models for original and noise images

Motohiro Tabuchi, Nobumoto Yamane

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Citations (Scopus)

Abstract

An adaptive Wiener filter for denoising X-ray CT image has been proposed based on the universal Gaussian mixture distribution model (UNI-GMM). In this method, the UNI-GMM is estimated by the statistical learning method using two sets of pair images, one of which is an observed (low dose) X-ray CT image set and the other is an original (high dose) X-ray CT image set. Owing to the physical limitations of CT scanners, the original (high dose) X-ray CT image also includes considerable noise that prevented precise learning of the UNI-GMM. On the other hand, the noise included in the X-ray CT images is the specific artifact which is called streak artifact and is known to be statistically non-stationary. In the previously proposed method, the artifact is treated to be stationary for simplicity. Thus the restored images include residual noise due to the non-stationary noise. In this paper, the UNI-GMM method is improved by a two stages product modeling. First, the UNI-GMM for the original image is estimated using a low noise natural image set that include scenes, portraits and still pictures, to prevent the effect of noise on the original (high dose) CT images. Second, the UNI-GMM for the noise image is estimated using a noise image set casted by subtracting the original X-ray CT images from the observed X-ray CT images. Simulation results show that the proposed product UNI-GMMs performs better than the conventional stationary noise model simply learned using X-ray CT images.

Original languageEnglish
Title of host publicationIEEE Region 10 Annual International Conference, Proceedings/TENCON
Pages1679-1684
Number of pages6
DOIs
Publication statusPublished - 2010
Event2010 IEEE Region 10 Conference, TENCON 2010 - Fukuoka, Japan
Duration: Nov 21 2010Nov 24 2010

Other

Other2010 IEEE Region 10 Conference, TENCON 2010
CountryJapan
CityFukuoka
Period11/21/1011/24/10

Fingerprint

X rays
Dosimetry
Adaptive filters

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Computer Science Applications

Cite this

Tabuchi, M., & Yamane, N. (2010). Denoising X-ray CT images based on product Gaussian mixture distribution models for original and noise images. In IEEE Region 10 Annual International Conference, Proceedings/TENCON (pp. 1679-1684). [5686039] https://doi.org/10.1109/TENCON.2010.5686039

Denoising X-ray CT images based on product Gaussian mixture distribution models for original and noise images. / Tabuchi, Motohiro; Yamane, Nobumoto.

IEEE Region 10 Annual International Conference, Proceedings/TENCON. 2010. p. 1679-1684 5686039.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Tabuchi, M & Yamane, N 2010, Denoising X-ray CT images based on product Gaussian mixture distribution models for original and noise images. in IEEE Region 10 Annual International Conference, Proceedings/TENCON., 5686039, pp. 1679-1684, 2010 IEEE Region 10 Conference, TENCON 2010, Fukuoka, Japan, 11/21/10. https://doi.org/10.1109/TENCON.2010.5686039
Tabuchi M, Yamane N. Denoising X-ray CT images based on product Gaussian mixture distribution models for original and noise images. In IEEE Region 10 Annual International Conference, Proceedings/TENCON. 2010. p. 1679-1684. 5686039 https://doi.org/10.1109/TENCON.2010.5686039
Tabuchi, Motohiro ; Yamane, Nobumoto. / Denoising X-ray CT images based on product Gaussian mixture distribution models for original and noise images. IEEE Region 10 Annual International Conference, Proceedings/TENCON. 2010. pp. 1679-1684
@inproceedings{3ddf9e217b344c4aa7206b03a66d443d,
title = "Denoising X-ray CT images based on product Gaussian mixture distribution models for original and noise images",
abstract = "An adaptive Wiener filter for denoising X-ray CT image has been proposed based on the universal Gaussian mixture distribution model (UNI-GMM). In this method, the UNI-GMM is estimated by the statistical learning method using two sets of pair images, one of which is an observed (low dose) X-ray CT image set and the other is an original (high dose) X-ray CT image set. Owing to the physical limitations of CT scanners, the original (high dose) X-ray CT image also includes considerable noise that prevented precise learning of the UNI-GMM. On the other hand, the noise included in the X-ray CT images is the specific artifact which is called streak artifact and is known to be statistically non-stationary. In the previously proposed method, the artifact is treated to be stationary for simplicity. Thus the restored images include residual noise due to the non-stationary noise. In this paper, the UNI-GMM method is improved by a two stages product modeling. First, the UNI-GMM for the original image is estimated using a low noise natural image set that include scenes, portraits and still pictures, to prevent the effect of noise on the original (high dose) CT images. Second, the UNI-GMM for the noise image is estimated using a noise image set casted by subtracting the original X-ray CT images from the observed X-ray CT images. Simulation results show that the proposed product UNI-GMMs performs better than the conventional stationary noise model simply learned using X-ray CT images.",
author = "Motohiro Tabuchi and Nobumoto Yamane",
year = "2010",
doi = "10.1109/TENCON.2010.5686039",
language = "English",
isbn = "9781424468904",
pages = "1679--1684",
booktitle = "IEEE Region 10 Annual International Conference, Proceedings/TENCON",

}

TY - GEN

T1 - Denoising X-ray CT images based on product Gaussian mixture distribution models for original and noise images

AU - Tabuchi, Motohiro

AU - Yamane, Nobumoto

PY - 2010

Y1 - 2010

N2 - An adaptive Wiener filter for denoising X-ray CT image has been proposed based on the universal Gaussian mixture distribution model (UNI-GMM). In this method, the UNI-GMM is estimated by the statistical learning method using two sets of pair images, one of which is an observed (low dose) X-ray CT image set and the other is an original (high dose) X-ray CT image set. Owing to the physical limitations of CT scanners, the original (high dose) X-ray CT image also includes considerable noise that prevented precise learning of the UNI-GMM. On the other hand, the noise included in the X-ray CT images is the specific artifact which is called streak artifact and is known to be statistically non-stationary. In the previously proposed method, the artifact is treated to be stationary for simplicity. Thus the restored images include residual noise due to the non-stationary noise. In this paper, the UNI-GMM method is improved by a two stages product modeling. First, the UNI-GMM for the original image is estimated using a low noise natural image set that include scenes, portraits and still pictures, to prevent the effect of noise on the original (high dose) CT images. Second, the UNI-GMM for the noise image is estimated using a noise image set casted by subtracting the original X-ray CT images from the observed X-ray CT images. Simulation results show that the proposed product UNI-GMMs performs better than the conventional stationary noise model simply learned using X-ray CT images.

AB - An adaptive Wiener filter for denoising X-ray CT image has been proposed based on the universal Gaussian mixture distribution model (UNI-GMM). In this method, the UNI-GMM is estimated by the statistical learning method using two sets of pair images, one of which is an observed (low dose) X-ray CT image set and the other is an original (high dose) X-ray CT image set. Owing to the physical limitations of CT scanners, the original (high dose) X-ray CT image also includes considerable noise that prevented precise learning of the UNI-GMM. On the other hand, the noise included in the X-ray CT images is the specific artifact which is called streak artifact and is known to be statistically non-stationary. In the previously proposed method, the artifact is treated to be stationary for simplicity. Thus the restored images include residual noise due to the non-stationary noise. In this paper, the UNI-GMM method is improved by a two stages product modeling. First, the UNI-GMM for the original image is estimated using a low noise natural image set that include scenes, portraits and still pictures, to prevent the effect of noise on the original (high dose) CT images. Second, the UNI-GMM for the noise image is estimated using a noise image set casted by subtracting the original X-ray CT images from the observed X-ray CT images. Simulation results show that the proposed product UNI-GMMs performs better than the conventional stationary noise model simply learned using X-ray CT images.

UR - http://www.scopus.com/inward/record.url?scp=79951607553&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79951607553&partnerID=8YFLogxK

U2 - 10.1109/TENCON.2010.5686039

DO - 10.1109/TENCON.2010.5686039

M3 - Conference contribution

SN - 9781424468904

SP - 1679

EP - 1684

BT - IEEE Region 10 Annual International Conference, Proceedings/TENCON

ER -