Optimization of block-matching and 3D filtering (BM3D) algorithm in brain SPECT imaging using fan beam collimator: Phantom study

The purpose of this study is to model and optimize the block-matching and 3D filtering (BM3D) algorithm and to evaluate its applicability in brain single-photon emission computed tomography (SPECT) images using a fan beam collimator. For quantitative evaluation of the noise level, the coefficient of...

Full description

Saved in:

Bibliographic Details
Published in:	Nuclear engineering and technology Vol. 54; no. 9; pp. 3403 - 3414
Main Authors:	Do, Yongho, Cho, Youngkwon, Kang, Seong-Hyeon, Lee, Youngjin
Format:	Journal Article
Language:	English
Published:	Elsevier B.V 01.09.2022 Elsevier 한국원자력학회
Subjects:	Block-matching and 3D filtering (BM3D) algorithm Brain single photon emission computed tomography (SPECT) Optimization of noise reduction algorithm Quantitative evaluation of image quality 원자력공학 Block-matching and 3D filtering (BM3D) algorithm Quantitative evaluation of image quality Brain single photon emission computed tomography (SPECT) Optimization of noise reduction algorithm
ISSN:	1738-5733, 2234-358X
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Description
Summary:	The purpose of this study is to model and optimize the block-matching and 3D filtering (BM3D) algorithm and to evaluate its applicability in brain single-photon emission computed tomography (SPECT) images using a fan beam collimator. For quantitative evaluation of the noise level, the coefficient of variation (COV) and contrast-to-noise ratio (CNR) were used, and finally, a no-reference-based evaluation parameter was used for optimization of the BM3D algorithm in the brain SPECT images. As a result, optimized results were derived when the sigma values of the BM3D algorithm were 0.15, 0.2, and 0.25 in brain SPECT images acquired for 5, 10, and 15 s, respectively. In addition, when the sigma value of the optimized BM3D algorithm was applied, superior results were obtained compared with conventional filtering methods. In particular, we confirmed that the COV and CNR of the images obtained using the BM3D algorithm were improved by 2.40 and 2.33 times, respectively, compared with the original image. In conclusion, the usefulness of the optimized BM3D algorithm in brain SPECT images using a fan beam collimator has been proven, and based on the results, it is expected that its application in various nuclear medicine examinations will be possible.
ISSN:	1738-5733 2234-358X
DOI:	10.1016/j.net.2022.04.008