Ten patients with a dental implant in the esthetic zone (regions 13 to 23) were included (Figures 1 and 2). Research was carried out in compliance with the Helsinki Declaration. Patients were part of a randomized controlled trial on esthetics; the study was approved by the Medical Ethic Board of the University Medical Center Groningen, University of Groningen (METC 2010.246) as well as that written informed consent was obtained from all patients. The CBCT scans were made with an iCAT 3D exam scanner (KaVo Dental GmbH, Biberach, Germany), which scanner was validated for measuring bone thickness by Fourie et al. [17]. The method error of this scanner is very small, i.e., 0.05 mm (95% CI 0.03 to 0.07). The standard used voxel size was 0.30 and FoV was 100 × 100 mm on the CBCT scans. Bone measurements at implants on the CBCT scans was done using 3D image diagnostic and treatment planning software (NobelClinician, version 2.1 (Nobel Biocare - Guided Surgery Center, Mechelen, Belgium). A novelty is that this program, regularly used preoperatively, was employed to measure the buccal bone thickness (in mm), after implant surgery. To allow for reproducible measurements, a CBCT imaging and software protocol was developed.
CBCT imaging and software protocol
Acquired CBCT Digital Imaging and Communications in Medicine (DICOM) datasets were transferred to a computer. The CBCT images were exported in DICOM multi-file format and imported into Maxilim, version 2.3 (Medicim, Sint-Niklass, Belgium). Maxilim is a medical image computing program assessing the patients head anatomy and is used for diagnostics and preoperative planning of maxillofacial surgery. The input information for Maxilim is a 3D dataset, often (CB)CT data. The DICOM files of all patients were set continuously on Hounsfield unit (HU) isovalue 280. The implant used was set on HU isovalue 130. With Multimodality Image Registration using Information Theory (MIRIT), which has an accuracy of a subvoxel, the exact position of the implant could be recognized, determined and implemented in the patients DICOM files [16]. The MIRIT procedure is based on recognizing image similarities. The degree of similarity between intensity patterns in two images is determined, and consequently, the recognized image is registered automatically into one coordinate system. Image similarities are broadly used in medical imaging to enhance diagnostics. In the software program NobelClinician, the patients DICOM files were opened with the same HU isovalue of 280. An extra research tool was added to this software program by the program makers, so that the DICOM file from Maxilim was recognized by this program and the exact position of the implant, as determined in Maxilim, could be aligned with a planning implant in NobelClinician. Due to the alignment of a planning implant (with a known configuration) and an actual inserted implant into one image, measurements could take place at the exact buccal midline of the implant (Figure 3). The display of the implant and surrounding structures was set on bone value, so that the outline of the bony structures could be seen and measured. The buccal bone measurements at midline of the implant were performed with the standard provided measurement tools in the software program of NobelClinician.
Measuring procedure
The implant and patient dataset were exactly aligned by the MIRIT method, so that the distance from the central axis of the implant to the outer contour of the buccal bone could be measured. Area of interest was the upper 5 mm section of the implant, beginning at the neck of the implant towards the apical direction. Exact dimensions along the implant axis of each implant configuration used in the study were provided by the manufacturer. Buccal bone measurements (in mm) were performed calculating the distance to the buccal bone outline minus the radius of the interior contour of the implant. These buccal bone measurements were done for 5 mm at each millimeter along the axis, beginning at the neck of the implant (Figure 4). Measurements were repeated twice (with time interval to prevent recollection) by two independent operators (HJAM and KWS, both dentists) in a random order. Flow diagram of the consecutive steps has been depicted in Figure 5.
Statistical analysis
Continuous variables were expressed as a mean with standard deviation. Interobserver and intraobserver variability was assessed using two-way mixed intraclass correlation coefficient single measures analysis [18]. All analyses were performed using SPSS software (version 20.0).