A comparative evaluation of Cone Beam Computed Tomography (CBCT) and Multi-Slice CT (MSCT). Part II: On 3D model accuracy
Introduction
Successful dental implant rehabilitation requires accurate pre-operative surgical planning. Crucial to its success is accurate positioning of the implant prosthesis in the jaw to avoid damaging vital nervous and vascular structures. To realize those objectives, contemporary planning techniques rely on three-dimensional (3D) surface model representations of the jaw bone obtained from computed tomography (CT) scans [1], [2]. The 3D surface models can be used to plan the implant procedure pre-operatively and also to create physical cast models of the jaws using stereolithography technology [3], [4]. Those models facilitate surgical Implant placement procedure in conjunction with appropriate pre-operative planning [5].
Cone Beam Computed Tomography (CBCT) technology has been used for pre-operative implant planning, evaluation of the jaws and preliminary assessment of the bone volume required for orthognathic surgery [6], [7], [8]. CBCT provides images of diagnostic quality with reasonably low radiation dose [9], [10]. CBCT is used to create different two-dimensional (2D) tomographic slices and projection images as well as 3D volume and surface reconstructions. The geometric accuracy of CBCT based on phantom and dry skulls with metallic markers samples is well established [11], [12], [13], [14]. Also, 2D tomographic slices and projection images reconstructions were found to be highly accurate in comparison with physical measurements [15], [16].
However, the accuracy of 3D surface models reconstructions from CBCT has not yet been thoroughly evaluated. The quality of those models is largely dependent on the scanner type, scanning parameters and reconstruction settings [17]. Recently, the accuracy of CBCT 3D surface and volume reconstructions based on linear cephalometric measurements has been estimated to be within 1–2 mm [18], [19]. However, the accuracy of the linear measurements in those studies is both observer and landmark dependent. A method that matches CBCT 3D model with a ‘reference’ 3D model using point-based registration algorithm was proposed and used to evaluate the total geometric accuracy of CBCT 3D surface models [6]. This method is completely automatic and observer-independent and it evaluates the entire bone surface. The objective of this study is to assess the accuracy of 3D surface model from five clinical CBCT scanners in comparison with 3D surface models obtained from multi-slice CT (MSCT). A high resolution laser surface scanner is used to create the reference ‘gold standard’ 3D model.
Section snippets
Materials and methods
One dry human mandible was obtained with approval from the Institute for Biomedical Research (BIOMED of University Hasselt, Belgium). The mandible was first scanned by a XC50 Cross Scanner® (Metris, Leuven, Belgium) with three laser planes, mounted on a Wenzel LH57 3D coordinate measurement machine (Fig. 1). The laser scanner allows obtaining surfaces in the form of a point cloud with an accuracy of 15 μm. The mandible was placed in a plastic container and embedded in water to provide some level
Results
Mean deviations of the CBCT and MSCT measurements from the gold standard are summarized in Table 2. All deviations were statistically significant at p = 0.05, however, most deviations were small that their relevance is limited. MSCT had the least mean deviation (0.137) while for CBCT it was between 0.165 and 0.386 mm. The matching fit level of each model was marked with a colour bar. The mandibular border, posterior margins and some teeth margins were the parts that did not fit well with the laser
Discussion
This study was conducted to assess and compare the geometric accuracy of 3D surface model reconstructions from five clinical dental CBCT scanners with that of a clinical MSCT system. The results demonstrate higher segmentation accuracy of MSCT compared to CBCT. This could be due to that MSCT has a higher inherent image contrast with a better contrast to noise ratio in comparison with CBCT systems currently available [21]. The superiority of MSCT is mainly due to higher X-ray throughput and
Conclusion
CBCT still has somewhat lower segmentation accuracy than MSCT. Yet, anatomical landmarks and models are produced in a reliably and clinically applicable way. The accuracy of the mandibular contours was lower than the body because of the halation defects. Considering these results with low radiation, short scanning time and good image quality, CBCT could be helpful for surgery including implant placement in the orofacial region.
References (29)
- et al.
Neurovascularisation of the anterior jaw bones revisited using high resolution magnetic resonance imaging
Oral Surg Oral Med Oral Pathol Oral Radiol Endod
(2007) - et al.
Assessment of bone segmentation quality of cone-beam CT versus multislice spiral CT: a pilot study
Oral Surg Oral Med Oral Pathol Oral Radiol Endod
(2006) - et al.
Major mandibular surgical procedures as an indication for intraoperative imaging
J Oral Maxillofac Surg
(2008) - et al.
Accuracy of measurements of mandibular anatomy in cone beam computed tomography images.
Oral Surg Oral Med Oral Pathol Oral Radiol Endod
(2007) - et al.
Accuracy of linear measurements from imaging plate and lateral cephalometric images derived from cone-beam computed tomography
Am J Orthod Dentofacial Orthop
(2007) - et al.
Geometric accuracy of a newly developed cone-beam device for maxillofacial imaging
Oral Surg Oral Med Oral Pathol Oral Radiol Endod
(2007) - et al.
Clinical accuracy of 3 different types of computed tomography-derived stereolithographic surgical guides in implant placement
J Oral Maxillofac Surg
(2009) - et al.
Image artifact in dental cone-beam CT
Oral Surg Oral Med Oral Pathol Oral Radiol Endod
(2006) - et al.
Geometric accuracy of digital volume tomography and conventional computed tomography
Br J Oral Maxillofac Surg
(2008) - et al.
Lingual foramina on the mandibular midline revisited: a macroanatomical study
Clin Anat
(2007)