Quantification of metal artifacts on cone beam computed tomography images. (15th December 2011)

• Record Type: Journal Article
• Authors: Pauwels, Ruben; Stamatakis, Harry; Bosmans, Hilde; Bogaerts, Ria; Jacobs, Reinhilde; Horner, Keith; Tsiklakis, Kostas
• Is Part Of: Clinical oral implants research. Volume 24(2013)Supplement A100
• Page Start: 94
• Page End: 99
• Publication Date: 2011-12-15
• Subjects: cone beam computed tomography -- image analysis -- metal artifacts -- titanium; Dental implants -- Research -- Periodicals; 617.69
• DOI: 10.1111/j.1600-0501.2011.02382.x
• Languages: English
• ISSNs: 0905-7161

Abstract

Abstract: Objectives: To quantify metal artifacts obtained from a wide range of cone beam computed tomography (CBCT) devices and exposure protocols, to compare their tolerance to metals of different densities, and to provide insights regarding the possible implementation of metal artifact analysis into a QC protocol for CBCT. Materials and methods: A customized polymethyl methacrylate (PMMA) phantom, containing titanium and lead rods, was fabricated. It was scanned on 13 CBCT devices and one multi‐slice computed tomography (MSCT) device, including high‐dose and low‐dose exposure protocols. Artifacts from the rods were assessed by two observers by measuring the standard deviation of voxel values in the vicinity of the rods, and normalizing this value to the percentage of the theoretical maximum standard deviation. Results: For CBCT datasets, artifact values ranged between 6.1% and 27.4% for titanium, and between 10.% and 43.7% for lead. Most CBCT devices performed worse than MSCT for titanium artifacts, but all of them performed better for lead artifacts. In general, no clear improvement of metal artifacts was seen for high‐dose protocols, although certain devices showed some artifact reduction for large FOV or high exposure protocols. Conclusions: Regions in the vicinity of the metal rods were moderately or gravely affected, particularly in the area between the rods. In practice, the CBCT user has very limited possibilities to reduce artifacts. Researchers and manufacturers need to combine their efforts in optimizing exposure factors and implementing metal artifact reduction algorithms.