The patients who were edentulous in one or in both jaws and seeking implant-supported prosthesis treatment in the Department of Prosthodontics at Peking University School and Hospital of Stomatology from December 2017 to June 2018 were included in this study. The inclusion and exclusion criteria are shown in Table 2. The study was reviewed and approved by the Institutional Review Board of Peking University School and Hospital of Stomatology. The study was registered in the Chinese Clinical Trial Registry (ChiCTR-ONC-17014159). This study was undertaken with the understanding and written informed consent of each individual participant and was conducted in accordance with the World Medical Association’s Declaration of Helsinki (Version, 2013).
2. Design and fabrication of the radiographic template
The study protocol is summarized in Fig. 1. For each patient, after conventional impressions were taken for both the maxilla and mandible, stone models were fabricated. Maxillomandibular relationship registration was performed and verified, and the models were mounted in an articulator using facebow transfer technique. The wax-up dentures (Fig. 2a) were tried in the patient’s mouth, and after clinically necessary corrections, the wax-up dentures were sent back to the laboratory for the production of radiographic templates.
The wax-up was digitized using a table-top scanner (D2000, 3shape, Copenhagen K, Denmark). A radiopaque polymethyl methacrylate (PMMA) blank (Organical® PMMA, Organical, Berlin, Germany) was used to mill the dentitions on a 5-axis CNC milling machine (Fig. 2b, Organical® Multi S, Organical, Berlin, Germany). A transparent self-cure resin base was built on the model (Fig. 2c and 2d). Finally, a registration template (Diagnostic Template, Organical®, Berlin, Germany) was bonded to the lingual side of the radiopaque dentition to complete the radiographic template (Fig. 2e and 2f).
3. CBCT scan and virtual implant planning
The radiographic template was placed in patient’s mouth to confirm its fit. A silicone index was made to further stabilize the template,. With the templates in the patient’s mouth, a CBCT scan (VGi, New Tom, Verona, Italy, voxel size 0.25 mm3, field of view 12 cm x 8 cm, voltage 110 kV, tube current 3.5 mA) was made (Fig. 3). The image data were exported in the DICOM format and exported to virtual planning software (Organical® Dental Implant, ODI 18.104.22.168, Organical, Berlin, Germany). The radiopaque dentitions and the alveolar bone could be viewed in the software (Fig. 4).
Virtual implant planning was conducted in the software in accordance with the prosthetic treatment protocol (Fig. 4). The software aligned the spatial coordinates of the radiographic templates with its system coordinates by identifying the zirconia beads in the diagnostic plate in the DICOM data (Fig. 5). The position information of the virtually designed implants was then transferred into coordinate values that could be identified by the milling software. The planning data were then exported in the initial graphics exchange specification (IGES) format.
4. Milling of surgical guides
The implant planning data were transferred into the milling software (Organical® Mill2, Organical, Berlin, Germany), and the radiographic guide with the diagnostic template was fixed on the 5-axis CNC machine (Organical® Multi S, Organical, Berlin, Germany). Slots for the guide sleeve of each implant were milled on the radiographic templates. The guide sleeves were precisely installed into the slots (Fig. 6). Thus, the radiographic template was transferred into an implant surgical guide.
5. Guided surgery
All the surgeries were performed by two experienced dentists (S.P and J.L) at the Department of prosthodontics of Peking University School and Hospital. Before surgery, the surgical guide and the silicon index were disinfected in 0.12% chlorhexidine for 30 minutes. The surgical guide was positioned on the edentulous jaw with the interocclusal silicon index to confirm proper seating.
After local anaesthesia, the surgical guide was either fixed on patient’s alveolar ridge by three lateral fixation pins or retained using fixation anchors through guide sleeves after the first twist drill. A punch drill was used to remove the mucosa on top of the alveolar ridge, and a flapless guided implant placement protocol was followed.
Based on the virtual planning, the correct combination of drill handles and guided instruments was used for osteotomy site preparation, and the implants were installed (Bone level implant or Tissue level implant, Institut Straumann AG, Switzerland. Fig. 7). Guided bone regeneration (GBR) procedure was performed for one implant with bone graft material (0.25g, Bio-Oss®, Geistlich, Switzerland) and collagen membrane (Bio-Guide®, Geistlich, Switzerland).
6. Deviation measurement
After implant placement, a second CBCT scan was made for each patient.The post-surgical CBCT data were imported into Mimics (Mimics 19.0, Materialise, Leuven, Belgium), and post-operative digital models in the Standard Tessellation Language (STL) format were generated from the DICOM data. A 1.25 mm layer of bone around the implant was removed using the Masks, Morphology, and Boolean function in the software. Finally, the data of the bone structure of the edentulous maxilla and mandible together with the isolated implants in the STL format were exported from Mimics software and imported into the virtual planning software, where the data were superimposed with the pre-surgical CBCT image that contained the virtually planned implants (Fig. 8).
The deviation between the virtually planned and actually placed implant positions was measured at the neck and apex of each implant.
Four parameters were defined, namely, the global deviation, horizontal deviation, depth deviation, and angular deviation. All parameters, except for the angular deviation, were measured both at implant neck and apex.
The global deviation was defined as the 3D distance between the centres of the neck (or apex) of the corresponding virtually planned and actually placed implants. To calculate the lateral deviation, a plane perpendicular to the longitudinal axis of the planned implant and through its coronal or apical centre was defined and set as the reference plane. The horizontal deviation was defined as the distance between the coronal (or apical) centre of the planned implant and the point of intersection of the longitudinal axis of the placed implant with the reference plane. The depth deviation was defined as the distance between the coronal (or apical) centre of the placed implant and the reference plane. The angular deviation was defined as the three-dimensional angle between the longitudinal axis of the planned and placed implants (Fig. 9).
To analyse the factors contributing to implant deviation, the mean global deviation in the maxillary cases and that in the mandibular cases were compared. Mean global deviation in cases with and without lateral fixation pins were also compared.
Mean difference of inter-implant distance
The distances between each pair of neighboring implants were measured in the virtual planning and the actual post-surgery CBCT scans. The mean differences between the virtual and actual inter-implant distance at implant neck and apex area were calculated, this will represent the random errors in the surgical template (Fig. 10).
7. Data analysis
The data were analysed descriptively using statistical software ( IBM SPSS Statistics, v20.0; IBM Corp, Chicago, IL, USA). To determine the contributing factors for the deviation in implant position, the deviation values between the upper and lower jaws as well as between cases using fixation pins for the surgical guide and cases not using fixation pins were compared using independent t-tests.