A feature refinement approach for statistical interior CT reconstruction

	A feature refinement approach for statistical interior CT reconstruction
	Hu, Zhanli; Zhang, Yunwan; Liu, Jianbo; Ma, Jianhua; Zheng, Hairong; Liang, Dong
刊名	PHYSICS IN MEDICINE AND BIOLOGY
	2016
英文摘要	Interior tomography is clinically desired to reduce the radiation dose rendered to patients. In this work, a new statistical interior tomography approach for computed tomography is proposed. The developed design focuses on taking into account the statistical nature of local projection data and recovering fine structures which are lost in the conventional total-variation (TV)-minimization reconstruction. The proposed method falls within the compressed sensing framework of TV minimization, which only assumes that the interior ROI is piecewise constant or polynomial and does not need any additional prior knowledge. To integrate the statistical distribution property of projection data, the objective function is built under the criteria of penalized weighed least-square (PWLS-TV). In the implementation of the proposed method, the interior projection extrapolation based FBP reconstruction is first used as the initial guess to mitigate truncation artifacts and also provide an extended field-of-view. Moreover, an interior feature refinement step, as an important processing operation is performed after each iteration of PWLS-TV to recover the desired structure information which is lost during the TV minimization. Here, a feature descriptor is specifically designed and employed to distinguish structure from noise and noise-like artifacts. A modified steepest descent algorithm is adopted to minimize the associated objective function. The proposed method is applied to both digital phantom and in vivo Micro-CT datasets, and compared to FBP, ART-TV and PWLS-TV. The reconstruction results demonstrate that the proposed method performs better than other conventional methods in suppressing noise, reducing truncated and streak artifacts, and preserving features. The proposed approach demonstrates its potential usefulness for feature preservation of interior tomography under truncated projection measurements.
收录类别	SCI
原文出处	http://iopscience.iop.org/article/10.1088/0031-9155/61/14/5311/meta
语种	英语
内容类型	期刊论文
源URL	[http://ir.siat.ac.cn:8080/handle/172644/10467]
专题	深圳先进技术研究院_医工所
作者单位	PHYSICS IN MEDICINE AND BIOLOGY
推荐引用方式 GB/T 7714	Hu, Zhanli,Zhang, Yunwan,Liu, Jianbo,et al. A feature refinement approach for statistical interior CT reconstruction[J]. PHYSICS IN MEDICINE AND BIOLOGY,2016.
APA	Hu, Zhanli,Zhang, Yunwan,Liu, Jianbo,Ma, Jianhua,Zheng, Hairong,&Liang, Dong.(2016).A feature refinement approach for statistical interior CT reconstruction.PHYSICS IN MEDICINE AND BIOLOGY.
MLA	Hu, Zhanli,et al."A feature refinement approach for statistical interior CT reconstruction".PHYSICS IN MEDICINE AND BIOLOGY (2016).