Metal artifact reduction in computed tomography based on prior interpolation

Ming, 李铭 Li; Yan-fei, 卢彦飞 Lu; Yuan, Gang; Zhong-yi, 吴中毅 Wu; Tao, 张涛 Zhang

doi:10.3788/yjyxs20153006.1032

Cited by 1 publication

(1 citation statement)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The current metal artifacts correction methods mainly include: the projection images interpolation correction method 3 – 5 , iterative reconstruction method 6 – 9 , prior image correction method 10 , dual-energy/energy spectrum computed tomography correction method 11 – 13 and correction method based on deep learning 14 – 18 . The projection images interpolation correction method uses the projection data of the surrounding normal tissue to perform linear interpolation on the metal area 3 , or normalizes the projection images by using the prior images and performs linear interpolation 4 , and obtain the correct images through back projection.…”

Section: Introductionmentioning

confidence: 99%

Metal artifacts reduction in kV-CT images with polymetallic dentures and complex metals based on MV-CBCT images in radiotherapy

Shi

Song

et al. 2023

Sci Rep

View full text Add to dashboard Cite

This paper proposes a metal artifact reduction method of using MV-CBCT images to correct metal artifacts in kV-CT images, especially for the complex metal artifacts caused by multi-metal interaction of patients with head and neck tumors. The different tissue regions are segmented in the MV-CBCT images to obtain template images and the metal region is segmented in the kV-CT images. Forward projection is performed to get sinogram of the template images, kV-CT images and metal region images. Artifact images can be reconstructed through those sonograms. Corrected images is generated by subtracting the artifact images from the original kV-CT images. After the first correction, the template images are generated again and brought into the previous step for iteration to get better correction result. CT data set of 7 patients are used in this study, compared with linear interpolation metal artifact (LIMAR) and normalized metal artifact reduction method, mean relative error of CT value is reduced by 50.5% and 63.3%, noise is reduced by 56.2% and 58.9%. The Identifiability Score of the tooth, upper/lower jaw, tongue, lips, masseter muscle and cavity in the corrected images by the proposed method have significantly improved (P < 0.05) than original images. The artifacts correction method proposed in this paper can effectively remove the metal artifacts in the images and greatly improve the CT value accuracy, especially in the case of multi-metal and complex metal implantation.

show abstract