TY - JOUR
T1 - Evaluation of artifacts generated by zirconium implants in cone-beam computed tomography images
AU - Vasconcelos, Taruska Ventorini
AU - Bechara, Boulos B.
AU - McMahan, Clyde Alex
AU - Freitas, Deborah Queiroz
AU - Noujeim, Marcel
N1 - Publisher Copyright:
© 2016 Elsevier Inc.
PY - 2017/2/1
Y1 - 2017/2/1
N2 - Objective To evaluate zirconium implant artifact production in cone beam computed tomography images obtained with different protocols. Study Design One zirconium implant was inserted in an edentulous mandible. Twenty scans were acquired with a ProMax 3D unit (Planmeca Oy, Helsinki, Finland), with acquisition settings ranging from 70 to 90 peak kilovoltage (kVp) and voxel sizes of 0.32 and 0.16 mm. A metal artifact reduction (MAR) tool was activated in half of the scans. An axial slice through the middle region of the implant was selected for each dataset. Gray values (mean ± standard deviation) were measured in two regions of interest, one close to and the other distant from the implant (control area). The contrast-to-noise ratio was also calculated. Results Standard deviation decreased with greater kVp and when the MAR tool was used. The contrast-to-noise ratio was significantly higher when the MAR tool was turned off, except for low resolution with kVp values above 80. Selection of the MAR tool and greater kVp resulted in an overall reduction of artifacts in images acquired with low resolution. Conclusions Although zirconium implants do produce image artifacts in cone-bean computed tomography scans, the setting that best controlled artifact generation by zirconium implants was 90 kVp at low resolution and with the MAR tool turned on.
AB - Objective To evaluate zirconium implant artifact production in cone beam computed tomography images obtained with different protocols. Study Design One zirconium implant was inserted in an edentulous mandible. Twenty scans were acquired with a ProMax 3D unit (Planmeca Oy, Helsinki, Finland), with acquisition settings ranging from 70 to 90 peak kilovoltage (kVp) and voxel sizes of 0.32 and 0.16 mm. A metal artifact reduction (MAR) tool was activated in half of the scans. An axial slice through the middle region of the implant was selected for each dataset. Gray values (mean ± standard deviation) were measured in two regions of interest, one close to and the other distant from the implant (control area). The contrast-to-noise ratio was also calculated. Results Standard deviation decreased with greater kVp and when the MAR tool was used. The contrast-to-noise ratio was significantly higher when the MAR tool was turned off, except for low resolution with kVp values above 80. Selection of the MAR tool and greater kVp resulted in an overall reduction of artifacts in images acquired with low resolution. Conclusions Although zirconium implants do produce image artifacts in cone-bean computed tomography scans, the setting that best controlled artifact generation by zirconium implants was 90 kVp at low resolution and with the MAR tool turned on.
UR - http://www.scopus.com/inward/record.url?scp=85009230797&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85009230797&partnerID=8YFLogxK
U2 - 10.1016/j.oooo.2016.10.021
DO - 10.1016/j.oooo.2016.10.021
M3 - Article
C2 - 28086998
AN - SCOPUS:85009230797
SN - 2212-4403
VL - 123
SP - 265
EP - 272
JO - Oral Surgery Oral Medicine and Oral Pathology
JF - Oral Surgery Oral Medicine and Oral Pathology
IS - 2
ER -