Abstract
The aim of the present study was to assess the clinical relevance of the potential mechanical error (intrinsic error) caused by the cylinder-burr gap in a ‘single type’ stereolithographic surgical template in implant guided surgery. 129 implants were inserted in 12 patients using 18 templates. The pre- and postoperative computed tomography (CT) scans were matched allowing comparison of the planned implants with the placed ones. Considering only the angular deviation values, the t test was used to determine the influence of the guide fixation and the arch of support on accuracy values. The Pearson correlation coefficient was used to correlate angular deviation and bone density. The intrinsic error was mathematically evaluated. t test results indicated that the use of fixing screws ( P = 009) and the upper arch support ( P = 027) resulted in better accuracy. The Pearson correlation coefficient (0.229) indicated a significant linear correlation between angular deviations and bone density ( P = 009). A mean intrinsic error of 2.57 was mathematically determined considering only the angular deviation, as it was not influenced by other variables. The intrinsic error is a significant factor compared to all the variables that could potentially affect the accuracy of computer-aided implant placement.
The use of CAD-CAM stereolithographic surgical guides, defined by Fortin et al. as semi-active systems for dental implant placement, is designed to provide improved control and eliminate the risks involved in standard implant surgery. The rapid development of this poorly documented technology has led to unrealistic clinical expectations for the efficacy and ease-of-use of this technology, while the risk of deviation (transfer error from the software-planning stage to the surgical field) remains substantial.
Computer-aided oral implant surgery involves a sequence of diagnostic and therapeutic events and errors can creep in at different stages. The described cumulative loss of accuracy is the sum of single errors throughout the ‘computer-aided implant placement cascade’.
Errors may arise from cone-beam computed tomography (CBCT) or multi-slice CT (MSCT) imaging, during different phases, including image acquisition and data processing, from software planning (conversion, segmentation, volume rendering and manual removal of artefacts) and from the transfer of the planning data.
An error might also occur during the manufacturing of the surgical template for example in the simulation software, the precision of the stereolithographic machine, production and quality control, rigidity and physical properties of the material used, the precision of the guide cylinders and metal tubes, and verification of the guide.
Inaccurate positioning of the guide in the mouth (flap intervention, improper or tilted seating, resilience of the anatomical structures that support the guide); improper guide fixation (angle, location, number of fixation screws); the presence of a rotational allowance of drills in tubes; the shape (straight or tapered) and sharpness of the drills; reduced mouth opening; and a non-guided insertion may cause a deviation between the postoperative position of the implants and the preoperative plan.
A lack of accuracy could be related to the technical procedure or depend on the hardware used. Errors during positioning of the surgical template are categorized as procedure-related and give rise to a decrease in accuracy, while the accuracy or rigidity of the surgical template is considered to be product-related. The significance of individual events in determining an accurate result has never been investigated.
It has not been determined which of the different steps of computer-aided implantology (CAI) may more frequently give rise to error. Based on the authors’ clinical experience, one of the potentially clinically relevant errors may be mechanical error caused by the bur-guide cylinder gap, definable as an ‘intrinsic error’ of the surgical guide. In the ‘multiple-type’ surgical guide, equipped with 5 mm long guiding cylinders with an inner diameter that is 0.15–0.20 mm larger than the respective bur, the tolerance theoretically allows a deviation angle of approximately 2.29° which, at a hypothetical distance of 20 mm from the cylinder, results in a lateral deviation of approximately 1 mm.
The aim of the present study was to determine the accuracy of a ‘single type’ stereolithographic surgical guide (External Hex Safe ® -Materialise Dental, Leuven, Belgium) and the relative importance of the error that arises from tolerance among the various components of this type of CAD-CAM surgical template (intrinsic error) and its incidence on the cumulative sum of all the errors throughout the computer-aided implant placement cascade.
The influence of the guide fixation (fixed vs not fixed), arch of support (upper vs lower arch) and bone density on intrinsic error was determined.
Materials and methods
A ‘single type’ stereolithographic surgical guide (External Hex Safe ® , Materialise Dental, Leuven, Belgium) was used for 12 patients (18 templates; 129 implants), partially (Kennedy class I) or totally edentate, who required an implant prosthetic rehabilitation ( Table 1 ). The patient’s average age was 55 years with a 3:1 sex ratio (male/female) ( Table 1 ). All patients consecutively treated with CAI between February 2004 and June 2011 were included in this retrospective study.
| Mean | Total | Number of guides | Number of implants | |
|---|---|---|---|---|
| Age | 55 | |||
| Number of implant | 129 | |||
| Number of subjects | 12 | 18 | 129 | |
| Gender | ||||
| Male | 9 | 13 | 102 | |
| Female | 3 | 5 | 27 | |
| Type of edentulism | ||||
| Total | 14 | 112 | ||
| Partial | 4 | 17 | ||
| Arch | ||||
| Maxilla | 10 | 78 | ||
| Mandible | 8 | 51 | ||
| Type of SAFE ® Guide | ||||
| Fixed | 10 | 75 | ||
| Not fixed | 8 | 54 | ||
| Surgical guide support | ||||
| Mucosa | 13 | 103 | ||
| Bone | 4 | 18 | ||
| Teeth | 1 | 8 | ||
| Surgical technique | ||||
| Flapless | 14 | 111 | ||
| Open-flap | 4 | 18 | ||
The surgical interventions were performed by the same operator (MC) who undertook the virtual surgical planning, using implant planning software (SimPlant ® , Materialise Dental, Leuven, Belgium). The protocol employed in this clinical study consisted of an integrated treatment sequence that involved the following steps. First, development of a radiopaque diagnostic template (scanno-guide) consisting in an exact replica of the removable prosthesis (partial or total) that achieved the patient’s aesthetic and functional requirements. Second, a CT scan of the patient’s arch was taken with a spiral CT device (Asteion Multi Toshiba Medical System, Rome, Italy). The scans included the scanno-guide. The CT parameters used were 0° gantry tilt, high resolution bone Kernel, 0.5 mm nominal slice thickness, 0.5 mm interval, and 0.5 mm pitch. Third, digital three dimensional (3D) CT-based surgical planning was undertaken. The computer program (SimPlant ® ) uses the original CT data, in DICOM format, to produce axial, 3D, panoramic and cross-sectional images, all of which are visible at the same time in four interactive windows on a computer monitor. With this software, implants are virtually placed according to bone anatomy and prosthetic design. The Hounsfield unit (HU) threshold used was Simplant’s predefined for bone (250–3071 HU). Fourth, CAD of the stereolithographic surgical guide was undertaken. The clinician designs the drilling template in the CAD environment ( Fig. 1 ).
Fifth, CAM of the stereolithographic surgical guide was undertaken to transfer the digital planning to the surgical environment. The External Hex Safe ® surgical guides were classified according to the type of supporting anatomical structure (teeth, bone, mucosa). The teeth-supported guides used were, in all cases, free ending templates and were seated and stabilized with the help of natural teeth ( Fig. 2 A) . The bone-supported guides ( Fig. 3 A) required an open flap reflection ( Fig. 3 B). The mucosa ( Fig. 4 A and B) and teeth-supported guides ( Fig. 2 B) permitted a flapless/transmucosal approach. This fully guided implant system allowed for controlled osteotomy site preparation ( Fig. 5 A) and implant placement in three dimensions ( Fig. 5 B). The system may be used with implants of different brands, with external hexagon and platform within 4.00 mm. Specific cylinders, so-called master tubes ( Fig. 6 A) (inner diameter 4.2 mm and height 4 mm) are embedded within the acrylic resin guide to accommodate mucotome ( Fig. 7 A) or implant mounting devices ( Fig. 7 D) that intimately engage the cylinders. To accommodate the drill handles, a tube adapter, called the internal tube ( Fig. 6 B, C) (height 5 mm and inner diameter 3.2 mm), is positioned within the master tube. After implant site preparation, the internal tube, which is 0.2 mm smaller than the master tube, is removed allowing close contact between the master tube and the implant mounting device. The first drill, in a flapless surgical technique (teeth or mucosa surgical guide), used to punch and remove gingival soft tissue ( Fig. 7 A) is a mucotome, with an outer diameter of 4.00 mm. In some cases, the templates were firmly fixed to the jaw (Group Safe ® -Fixed) using at least three fixation screws ( Fig. 8 A–E) . An inter-occlusal putty index was used to confirm proper seating, before fixing the template. In other cases, the surgical template was manually held in place by the surgeon (Group Safe ® -not Fixed) ( Fig. 8 F–L).
The sixth step is computer-aided surgery. 129 implants (Plan 1 Health, Amaro, Udine, Italy), cylindrical, with an external hexagon (diameter ranging from 3.75 to 4.00 mm and length 10–18 mm) were inserted, in partially and completely edentulous patients, using stereolithographic templates. Osteotomy site-specific drills, that have vertical stops to control apico-coronal site preparation, were used. Only two size types of single-use drills with physical stops were used: pilot drill (diameter 2.8 mm (top)/2.0 mm (bottom)) ( Fig. 7 B) and final drill (diameter 3.00 mm (top)/3.15 mm (bottom)) ( Fig. 7 C). Countersinking was not performed. Implant placement was performed using specific delivery mounts (implant holder length 4–15 mm, diameter 4.00 mm) to a controlled angulation and apico-coronal depth ( Fig. 9 ).
In step seven, as described by D’Haese et al. a postoperative CT was undertaken, using the same preoperative CT parameters, by all patients and an iterative closest point (ICP) algorithm was used to match the jaw of the preoperative CT with that of the postoperative CT (the software runs until it finds the best overlap between the images of pre- and postoperative jaws) ( Fig. 8 E, L). This allows comparison of the planned implants with the placed ones and the determination and calculation of four parameter deviations (i.e. global apical and coronal, depth, lateral and angular deviation) by using their 3D coordinates at the apical and coronal level ( Fig. 10 ). The average bone density of the implant site was measured in HU using a specific tool of the Simplant ® software.
Step eight was the evaluation of the intrinsic error. In order to assess the importance of intrinsic error in determining any discrepancy between the planned and the final position of the inserted implant, only the angular deviation was considered. To determine the intrinsic error of the guide mathematically, the tolerance between the master tube and internal tube of the guide ( Fig. 6 ) and between the internal tube and the drills ( Fig. 11 A, B ) were evaluated.
Statistical analysis
Data was evaluated using SPSS ® software (Statistical Package for Social Science, IBM Corporation, NY, USA). The quantitative data for each group was described with frequency distribution, mean values, standard deviations and median values.
Considering only the angular deviation values (angular deviation values are not influenced by other variables such as the distance between the bottom of the tube and the alveolar ridge and the implant length) the t test was used to determine the influence of the use of the different types of single-type guides (fixed vs not-fixed) and arch of support (upper vs lower arch) on accuracy values. The threshold for significance was set at P ≤ .05.
Correlations between angular deviation and bone density were tested with the Pearson correlation coefficient. The threshold for significance was set at P ≤ .01.
Results
Accuracy
12 adults were included in this study. 6 patients were treated in both arches, resulting in 129 planned and inserted implants. Patient and treatment characteristics are summarized in Table 1 .
In Group Safe ® -Fixed the number of CAI interventions was 10, (75 planned and inserted implants), and in Group Safe ® -not Fixed, 8 CAI interventions (54 planned and inserted implants) were undertaken. In Group Safe ® -Fixed and Group Safe ® -not Fixed ( Table 2 ) the global (coronal and apical), angular, depth and lateral deviations were determined via the image registration technique ( Fig. 10 ).
| Safe ® System | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Fixed | Not fixed | |||||||||
| Mean | Maximum | Minimum | Standard deviation | Count | Mean | Maximum | Minimum | Standard deviation | Count | |
| Coronal deviation (mm) | 1.59 | 3.63 | .13 | .68 | 75 | 1.55 | 2.79 | .31 | .59 | 54 |
| Angle deviation (°) | 4.11 | 14.34 | .28 | 2.40 | 75 | 5.46 | 15.25 | .10 | 3.38 | 54 |
| Depth deviation (mm) | .98 | 3.45 | .03 | .74 | 75 | .63 | 1.58 | .05 | .43 | 54 |
| Apical deviation (mm) | 2.07 | 4.47 | .44 | .88 | 75 | 2.05 | 4.23 | .34 | .89 | 54 |
| Lateral deviation (mm) | 1.06 | 2.57 | .12 | .63 | 75 | 1.36 | 2.61 | .29 | .58 | 54 |
The mean angular deviation between planned and placed implants in Group Safe ® -Fixed was 4.11° (range 14.34–0.28; SD 2.40) ( Table 2 ). In Group Safe ® -not Fixed (54 implants; 8 templates) the mean angular deviation was 5.46° (range 15.25–0.10; SD 3.38) ( Table 2 ).
Results of the t test with regards to the influence of the variables (arch and surgical technique) on the angular deviation parameter, appear to show that these variables exerted a significant effect on the accuracy of the surgery ( Table 3 ).
