Influence of image‐viewers and artifacts on implant length measurements in cone‐beam computed tomography: an in vitro study

Abstract

This preclinical in vitro study compared the accuracy of implant lengths measured in two different image‐viewers, and examined whether implant‐induced artifacts affected the implant length measurements on CBCT images. A resin edentulous mandibular model, with multiple adjacent implants in the posterior segments, was acquired with a CBCT machine. In two different image‐viewers, two observers independently measured the implant length. Vertical measurements on CBCT images were carried out twice at each session, and repeated one week later. The results demonstrated no significant differences between actual and measured implant lengths. The differences in the mean error for vertical measurements using the two different image‐viewers (cross‐sectional images: OsiriX viewer = −0.01 ± 0.03 mm, NewTom viewer = −0.05 ± 0.09 mm, p‐value = 0.056; sagittal images: OsiriX viewer = −0.03 ± 0.04 mm; NewTom viewer = −0.04 ± 0.10 mm, p‐value = 0.24) were not statistically significant. This in vitro investigation suggests that the accuracy of implant length measurements on CBCT images was not influenced by image‐viewers or by the presence of implant‐induced artifacts. The presence of multiple adjacent implants in the posterior segments of the mandible is not likely to impact the measurements made between the implant apex and vital structures on CBCT images.

Mesiodistal Measurements for Dental Implant Planning: A Prospective Clinical Study of Linear Measurements on Cone-Beam Computed Tomography Images in Comparison with Caliper-Based Measurements on Plaster Casts

“…Other potential limitations include artefacts related to the presence of titanium implants and the high-resolution CBCT acquisitions. The FOV size and acquisition mode are the settings utilized in our earlier preclinical study [ 14 ]. In the present study, 3/27 CBCT acquisitions presented lower quality images.…”

“…Visible motion artefacts, probably related to the high-resolution protocol requiring longer scan times, making images susceptible to patient movement during acquisition [ 21 ], could have contributed to the maximum 0.6 mm overestimation on CBCT views. Metal and titanium implants produce streak artefacts and beam-hardening artefacts [ 14 , 21 ]. We carefully selected the landmarks (the center of each dental implant) with the on-screen cursor (guided by a computer mouse).…”

“…The authors reported that OsiriX viewer was very reliable for measurement purposes (less than 0.1 mm difference between real and OsiriX measurements). In an in vitro preclinical study of vertical measurements on CBCT views acquired on the NewTom VGi CBCT machine used in the present study, we obtained slightly more accurate measurements on OsiriX viewer than on the NewTom viewer (Quantitative Radiology, Verona, Italy) [ 14 ]. The mean measurement differences were −0.01 ± 0.03 mm on cross-sectional and 0.03 ± 0.04 mm on sagittal views.…”

“…Many implant-related CBCT publications are presurgical CBCT studies [ 3 ] focused on bone height and the width of the edentulous ridge, with the issue of measurement accuracy on CBCT images receiving considerable attention over the years. Phantom heads [ 4 ]; dry skulls and human cadavers [ 2 , 5 , 6 , 7 ]; porcine mandibles [ 8 , 9 ]; beagle dog mandibles [ 10 ]; virtual models [ 11 , 12 ]; and reference objects, such as implants [ 13 , 14 ], have been used to assess the accuracy of linear measurements on CBCT views. Some authors have stated that CBCT views enable accurate measurements [ 6 ], whereas other authors report underestimated [ 5 ] or overestimated CBCT measurements [ 8 ].…”