Premium
WE‐AB‐207A‐12: HLCC Based Quantitative Evaluation Method of Image Artifact in Dental CBCT
Author(s) -
Chen Y,
Wu S,
Qi H,
Xu Y,
Zhou L
Publication year - 2016
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.4957765
Subject(s) - artifact (error) , cone beam computed tomography , projection (relational algebra) , computer vision , artificial intelligence , mathematics , iterative reconstruction , computer science , distortion (music) , algorithm , computed tomography , medicine , radiology , amplifier , computer network , bandwidth (computing)
Purpose: Image artifacts are usually evaluated qualitatively via visual observation of the reconstructed images, which is susceptible to subjective factors due to the lack of an objective evaluation criterion. In this work, we propose a Helgason‐Ludwig consistency condition (HLCC) based evaluation method to quantify the severity level of different image artifacts in dental CBCT. Methods: Our evaluation method consists of four step: 1) Acquire Cone beam CT(CBCT) projection; 2) Convert 3D CBCT projection to fan‐beam projection by extracting its central plane projection; 3) Convert fan‐beam projection to parallel‐beam projection utilizing sinogram‐based rebinning algorithm or detail‐based rebinning algorithm; 4) Obtain HLCC profile by integrating parallel‐beam projection per view and calculate wave percentage and variance of the HLCC profile, which can be used to describe the severity level of image artifacts. Results: Several sets of dental CBCT projections containing only one type of artifact (i.e. geometry, scatter, beam hardening, lag and noise artifact), were simulated using gDRR, a GPU tool developed for efficient, accurate, and realistic simulation of CBCT Projections. These simulated CBCT projections were used to test our proposed method. HLCC profile wave percentage and variance induced by geometry distortion are about 3∼21 times and 16∼393 times as large as that of the artifact‐free projection, respectively. The increase factor of wave percentage and variance are 6 and133 times for beam hardening, 19 and 1184 times for scatter, and 4 and16 times for lag artifacts, respectively. In contrast, for noisy projection the wave percentage, variance and inconsistency level are almost the same with those of the noise‐free one. Conclusion: We have proposed a quantitative evaluation method of image artifact based on HLCC theory. According to our simulation results, the severity of different artifact types is found to be in a following order: Scatter>Geometry>Beam hardening>Lag>Noise>Artifact‐free in dental CBCT