Premium
Fast CT metal artefacts correction based on derivative and region‐based filling
Author(s) -
Li YuanJin,
Chen Yang,
Luo LiMin,
Zhang PengCheng,
Zhang Quan
Publication year - 2011
Publication title -
journal of medical imaging and radiation oncology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.31
H-Index - 43
eISSN - 1754-9485
pISSN - 1754-9477
DOI - 10.1111/j.1754-9485.2011.02312.x
Subject(s) - computer vision , image quality , artificial intelligence , projection (relational algebra) , medicine , segmentation , computer science , image (mathematics) , algorithm
Metal artefacts seriously degrade the quality of the CT images. Blurring around the junctions between metal and non‐metal regions in CT images, metal artefacts often prevent right diagnoses, and even lead to misdiagnoses of patients. The aim of the study was to devise a fast and robust method to improve the quality of the artefact‐contaminated CT images. Methods: The proposed artefact correction includes the following five steps: metal object segmentation, forward projection, region‐based filling, adaptive scaling and final image reconstruction. Results: The feasibility of the proposed method in correcting metal artefacts was validated by experiments on both simulated and clinical images. Experiments showed the proposed correction could lead to fast and effective reduction of metal artefacts in CT images. Conclusions: Compared with other methods, the proposed method has less computational cost and allows a feasible and easy implantation into current CT imaging systems.