Introduction

Advanced dental technology have largely facilitated implant planning. The widespread availability of safer imaging systems, combined with potential risks of unplanned surgery, makes cone beam computed tomography (CBCT) imaging indispensable in implant dentistry [1-4]. A vital structure of importance when planning implants in the mandible is the mandibular canal (MC). Its sig-­ni­ficance arises from the risk of damaging the infe­rior alveolar neurovascular bundle [5-7]. Not only are the canal’s contents critical structures, but variations among patients in canal anatomy necessitate careful case-by-case assessment [8]. Such scrutiny is essential to achieve optimal conditions for viewing pre-implant images.

Several studies [9-21] have shown that MC visibility is affected by the imaging modality (2D vs. 3D), voxel size, tomographic plane angulation, monitor resolution, and radiation dose. Weissheimer et al. [22] reported that the accuracy of different imaging software systems varied when evaluating the upper airway. Similarly, Caetano et al. [23] found that the detection of vertical root fractures was influenced by the software used. These findings suggest that software choice may affect imaging tasks.

To our knowledge, it is unclear whether MC visibility is adversely affected by a third-party software or whether clinicians need to use a default software of their CBCT machines. This study aimed to address this question.

Material and methods

Patients

In this retrospective study, CBCT images of 300 lower molar implant sites were selected from patients who visited our dental radiology clinic between January 2021 and September 2023. All patients underwent CBCT examination for implant planning. Cases with artifacts (motion or metallic) or any pathology affecting MC visibility were excluded. This study was approved by the University Review Board (approval number: 20230491).

CBCT examination

CBCT scans were acquired using KODAK 9500 cone beam 3D system (Carestream, Rochester, NY, USA) with a flat panel detector. The imaging volume was cylindrical, with a height of 15-20.6 cm and a diameter of 9-18 cm, providing isotropic cubic voxels of 0.2-0.3 mm. Only scans obtained with a voxel size of 0.2 mm were included. Exposure parameters were 90 kV, 10 mA, and 10.8 s. Scans were acquired with 360° rotation in the occlusal position, with patients standing and their teeth in occlusion.

Image analysis

Two software programs were employed, including the original acquisition software (Kodak CS 3D v. 3.2.9) and a third-party software i.e., InVivoDental v. 5.0 (Anatomage Inc., San Jose, CA, USA). Two calibrated observers (R.B. and J.A.) with five years of CBCT expe­rience assessed MC visibility independently.

In Kodak software (Figure 1), the curved slicing module created pseudo-panoramic images. The arch was manually outlined on horizontal section, and the horizontal section was automatically displayed. The focal trough thickness was adjusted for each mandible, and only the mandible was displayed using the ‘region of interest’ function. Cross-sectional slices were generated by moving a vertical blue line perpendicular to the occlusal plane at the proposed implant site. Three slices per site were generated with a 2-mm interslice distance, corresponding to a 4-mm implant diameter.

In InVivo software (Figure 2), the arch section mo­dule was used similarly. A vertical green line was moved perpendicular to the occlusal plane at the proposed implant site, with a 2-mm cross-sectional interval. Images were zoomed to match the Kodak views.

MC visibility was scored as follows: 0 = not detected on any slice (poor), 1 = visible on 1 slice (good), 2 = visible on 2 slices (very good), 3 = visible on all 3 slices (excellent). The observers evaluated images on LCD monitors with the installed software, without altering window settings or applying image enhancements.

Statistical analysis

Interobserver agreement was assessed using intraclass correlation coefficients (ICCs) as follows: < 0.40 = poor, 0.40-0.60 = fair, 0.61-0.80 = good, > 0.80 = excellent. The Wilcoxon signed-rank test compared MC visibility between software programs. Statistical significance was set at p < 0.05. Analyses were performed using SPSS v. 16.0 (IBM SPSS Statistics, Chicago, IL, USA).

Results

Interobserver agreement for MC visibility was excellent: ICCs were 0.81 for Kodak and 0.99 for InVivo. Disagreements occurred in 31 cases for Kodak and 3 cases for InVivo, resolved by consensus.

The Wilcoxon signed-rank testing revealed a signi­ficant difference (p < 0.05). Visibility scores were identical in 269 implant sites. Kodak scored higher in 26 sites, while InVivo scored higher in 5 sites (Table 1).

Discussion

This study aimed to determine whether the visibi­lity of MC in cross-sectional CBCT images is influenced by the imaging software used for evaluation. Our results revealed a statistically significant difference, with Kodak performing marginally better than InVivo (26 vs. 5 sites). Although scores were identical in the majority of sites (269/300), this suggests that native software may provide a slight advantage. Therefore, while both programs are generally comparable, clinicians should be aware that software-related differences can occur, and should maintain consistency in software selection during implant planning.

Several studies have investigated the factors affecting MC visibility on CBCT images [15-17, 21, 24, 25], showing that visibility depends on slice angulation, location, dose, age, and gender. The visibility of MC declines in the mental area, as demonstrated in prior research [4, 16, 17], which may explain the excellent visibility observed in the current study, since only molar implant sites were included.

Our results align with previous studies, indicating that imaging software can impact the visibility of anatomical structures [22, 23, 26]. Weissheimer et al. [22] and Lo Giudice et al. [26] demonstrated variability in airway volume depending on the software used. Moreover, Caetano et al. [23] showed that even the detection of vertical root fractures is software-dependent. Our study confirms that software choice can also affect the visibility of crucial anatomical structures, such as the MC. Although the third-party software (InVivo­Dental) enhanced MC visibility at 5 implant sites, it may be due to differences in image processing or reconstruction, or enhanced image display. This also explains the higher interobserver agreement obtained with InVivo software (ICCs were 0.99 and 0.81 for InVivo and Kodak, respectively). In contrast, a few authors have evaluated the effect of imaging software on the accuracy of linear measurements performed on dry mandibles [27-30], and found that linear bone measurements were not influenced by the software used. These results are not directly comparable to the current study’s findings, as different software programs and study methodologies were employed.

Limitations should be noted. First, the retrospective design and restriction to a single CBCT device and two software versions limit generalizability. Second, the observers were not allowed to alter window/level or apply filters, yet default display curves of Kodak and InVivo were not disclosed and may differ. Third, important cli­nical and anatomical variables (i.e., molar region, cortical integrity, canal diameter, bone pattern, slice angulation) were not analyzed. Additionally, real-world clinical outcomes related to implant success and complication rates should be investigated to further understand the practical implications of the choice of software.

Conclusions

Although both Kodak CS 3D and InVivoDental software produced largely equivalent MC visibility scores, Kodak showed a statistically significant advantage in a minority of cases. This suggests that native acquisition software may offer a slight benefit, while third-party software generally performs comparably. Clinicians should be aware of potential software-related differences when planning implants and strive for consistency in software selection. Broader multicenter studies across different CBCT units, software versions, and clinical scenarios are recommended to validate and expand upon our findings.

Disclosures

Author contributions: Conceptualization: M.A.; Metho­dology: M.A., M.H.; Questionnaire adaptation: Not applicable; Investigation and data collection: R.B., J.A.; Formal analysis: M.A.; Data curation: M.A.; Writing of original draft: M.A., H.S.; Writing – review and editing: M.A., M.H., H.S.; Supervision: M.A.; Project administration: M.A. All authors have read and agreed to the published version of the manuscript.

Funding: This research received no external funding.

Institutional Review Board statement: This study was approved by the Review Board of the Jordan University of Science and Technology (approval number: 20230491 issued on 13 July 2023).

Informed consent statement: Not applicable.

Data availability statement: The data presented in this study are available within the article. Additional information is available from the corresponding author and can be provided upon reasonable request.

Acknowledgments: None.

Conflicts of interest: The authors declare no conflicts of interest.

AI use statement: No artificial intelligence tools were used in the preparation of this manuscript.

References