Several prior studies have compared the accuracy of navigation systems, static guide surgery, and freehand surgery in vitro [2, 13], with navigation systems yielding comparable or superior accuracy relative to static guide surgery and superior accuracy to that associated with freehand dental implant surgery. Similar findings have also been reported in clinical practice [14–16]. Dynamic navigation is thus a reliable clinical approach , the accuracy of which is not affected by surgeon experience level [15, 16].
TSFE was first introduced in the 1980s and was later modified by Summers , who introduced a technique in which a conventional drill was used to reach approximately 2 mm below the maxillary sinus floor. Hand osteotomes were then used to elevate the Schneiderian membrane after the infracture of the sinus floor . The residual bone height of 5–9 mm achieved good postoperative outcomes and minimal complications [20, 21]. The piezoelectric internal sinus elevation (PISE) technique was introduced in 2003 , and relies upon the use of a piezoelectric surgical tip in place of osteotomes. However, bone graft condensation is still required to elevate the sinus membrane in this procedure. For the present study, we used a drill to reach the planned position (within 1 mm of the bottom of the maxillary sinus floor) under the guidance of the dynamic navigation system. Then, a piezoelectric surgical tip was used to break through the sinus floor. After penetrating the sinus floor, osteotomes were inserted to elevate the sinus mucosa. Following membrane elevation, the graft was compacted into the osteotomy site and the implant was placed. Through the application of a piezoelectric device during this procedure, the duration and magnitude of osteotome use were dramatically reduced. Significant maxillary sinus perforation in this operative context may be avoided as evidenced by the immediate postoperative CBCT images.
In a recent systematic review of the accuracy of dynamic computer-aided implant placement conducted by Jorba-García et al. , the authors reported an average angular deviation of 3.68 degrees (95% CI: 3.61 to 3.74), an average coronal global deviation of 1.03 mm (95% CI: 1.01 to 1.04), an average apical global deviation of 1.34 mm (95% CI: 1.32 to 1.36) mm, an average lateral (2D) entry of 0.69 mm (95% CI: 0.67 to 0.72), and an average lateral (2D) apex of 0.9 mm (95% CI: 0.83 to 0.97) in clinical studies. Aydemir and Arısan  conducted dynamic navigation and freehand patient implant placement in the posterior maxilla, and measured an average coronal deviation of 1.01 mm (SD: 0.07 mm), an average apical deviation of 1.83 mm (SD: 0.12 mm), and an average angular deviation of 5.59 degrees (SD: 0.39 degrees) in the navigation group. In their study of 219 implants placed using a fully guided dynamic navigation approach, Block et al.  reported an average angular deviation of 2.97 degrees (SD: 2.09), an average coronal global deviation of 1.16 mm (SD: 0.59), an average apical global deviation of 1.29 mm (SD: 0.65), an average lateral (2D) entry of 0.74 mm (SD: 0.43), and an average lateral (2D) apex of 0.9 mm (SD: 0.55). Herein, we measured AD, EPHD, and APHD deviations between planned and actual implant placement, and found these values to be 3.656 ± 1.665 degrees, 1.073 ± 0.686 mm, and 1.086 ± 0.667 mm, respectively. Additionally, we found the angular deviation to be smaller for premolar sites relative to molar sites, and found that surgeon experience level had no impact on the overall accuracy of this approach.
Relative to dynamic navigation, the accuracy of freehand implant placement approaches is generally reported to be substantially reduced. For example, Block et al.  reported freehand placement to be associated with an average angular deviation of 6.5 degrees (SD: 4.21), an average coronal global deviation of 1.78 mm (SD: 0.77), an average apical global deviation of 2.27 mm (SD: 1.02), an average lateral (2D) entry of 1.19 mm (SD: 0.68), and an average lateral (2D) apex of 1.84 mm (SD: 1.05) when assessing 122 implants placed via a such an approach. Varga et al.  reported an average angular deviation of 7.13 degrees, an average coronal global deviation of 1.76 mm, and an average apical global deviation of 2.42 mm in the maxilla when conducting freehand surgery. Block reported that for 20 patients who underwent freehand surgical placement performed by two surgeons, mean angular deviation, platform lateral deviation, and apical lateral deviation values were 7.69 degrees, 1.15 mm, and 2.21 mm, respectively . In a separate study comparing planned and actual implant placement in a mental navigation group, Vercruyssen et al.  reported an average coronal deviation of 2.77 mm (SD:1.54 mm), an average apical deviation of 2.91 mm (SD: 1.52 mm), and an average angular deviation of 9.92 degrees (SD: 6.01 degrees). Aydemir and Arısan  compared dynamic navigation and freehand approaches in patients with bilateral edentulism in the posterior maxilla in whom sufficient bone volume was available to insert a standard implant (3.5-mm diameter and 10-mm long), reporting an average coronal deviation of 1.70 mm (SD: 0.13 mm), an average apical deviation of 2.51 mm (SD: 0.21 mm), and an average angular deviation of 10.04 degrees (SD: 0.83 degrees) in the freehand group. These results thus suggest that freehand implant placement is less accurate as compared to computer-aided approaches. This is a particularly important consideration when operating on an anatomical site that requires absolute accuracy during the implant placement procedure.
Pozzi and Moy evaluated the placement of 136 implants in 66 patients using a computer-guided template to perform flapless transcrestal maxillary sinus floor elevation with an expanding-condensing osteotomes protocol. This approach was able to achieve high rates of implant success when implanting implants into the posterior maxilla in a site with a single missing tooth with sufficient bone height (5–9 mm) .
The conventional TSFE procedure was used to estimate the position of the drill based on the preoperative CBCT images, positioning it to 1 mm below the floor of the maxillary sinus . For less experienced surgeons and even experienced surgeons dealing with the unique anatomical structure of the maxillary sinus, the actual direction and depth of the drill often differ from those planned positions. This can result in two outcomes. For one, this can cause damage to or perforation of the maxillary sinus mucosa. Second, it may result in the presence of more residual bone below the sinus floor than expected, affecting the subsequent use of hand osteotomes. If a piezoelectric device is used, however, it will prolong the operative duration. Dynamic navigation has noteworthy advantages when conducting posterior maxillary implant surgery with TSFE, particularly for premolars with a mesiodistal sloping maxillary sinus floor and for molars with a buccolingual sloping maxillary sinus floor (Fig. 3a, b). A dome-shaped elevator was used after piezoelectric surgery and before osteotome use for patients in whom such sloping was observed (Fig. 3c). The preoperatively planned apical implant position was 1 mm or less from the maxillary sinus floor, and we were able to use the visual dynamic navigation system to accurately drill to the lowest point of the sloping maxillary sinus floor. We then removed the bone upward along the lowest point of the maxillary sinus floor via piezoelectric osteotomy to achieve TSFE. We were then able to insert the traceable implant into the optimal position using the screen of the dynamic navigation system.
Several factors have the potential to impact the accuracy of dynamic navigation. CBCT image quality can be impacted by hardware, software, and human factors . Preoperative and postoperative CBCT imaging should be conducted using the same settings, as differing CBCT images would impact the accuracy of overlap between planned and actual implant positions. Patients have a registration device placed in the surgical site using silicone elastomer prior to undergoing CBCT scanning. Registration device movement would result in the incorrect positioning of the radiological fiducial markers. As per manufacturer recommendations, it is necessary to return the device to the company for calibration after every 50 uses. When not checked and repaired in a timely fashion, the system will yield poor accuracy. Owing to a lack of teeth contour and poor periodontal conditions, unstable registration and fixing device positioning can also result in inaccurate dynamic navigation during surgery. Intraoperative factors can also contribute to overall accuracy. Surgeons face a learning curve to achieve proficiency with this approach . The degree of mouth opening can also affect the maxillary posterior implant procedure, as it can confine the handpiece and drills to a limited area. During the TSFE operation, the density of residual bone and cortical bone at the base of the maxillary sinus can also impact operative accuracy. The nonuniform or discontinuous removal of bone from the floor of the maxillary sinus can result in inaccurate implant positioning, particularly for maxillary posterior sites with less residual bone height.
We have several suggestions that may aid others in their efforts to achieve optimal accuracy when conducting posterior maxillary implant surgery with TSFE using dynamic navigation. First, all patients should be fitted with a registration device using silicone elastomer in the implant site prior to CBCT imaging. The selection of the registration device type, placement, and finishing all warrant careful consideration. We were able to use a long registration device and additional silicone elastomer to extend the overall length of this device for patients in whom the free ends of the posterior teeth were missing. Patients can additionally bite rolled cotton on the registration during CBCT scanning. All of these steps can improve registration device stability and the repeatability of associated manipulation. During implant surgery, it is important to ensure the stability of the registration and fixation devices. We applied steady pressure to the registration device and selected evenly distributed registration points to complete the registration process. We were then able to perform infiltration anesthesia after registration, particularly in patients in whom the free ends of posterior teeth were missing, thereby avoiding the influence of soft tissue changes on registration accuracy. Stability during all stages of the implant placement procedure is critical, and the navigation guidance should be carefully followed, with the actual position of the implant in the surgical site being assessed to establish whether further modifications to the implant design are required. The implant should also be inserted under navigation guidance to ensure consistency between the planned and actual positioning of the implant.
As this was a retrospective study with a small sample size performed by two surgeons, additional large-scale prospective analyses will be essential to confirm the accuracy of dynamic navigation systems in the context of dental implant surgery with TSFE. Such future prospective randomized controlled trials should enroll surgeons with differing experience levels in order to gain further insight regarding the relationships among maxillary sinus anatomy, dentist seniority, and other variables associated with safe and accurate patient outcomes following implant surgery with TSFE.