Accuracy of Dynamic and Static Navigated Endodontic Microsurgery Using Single Vertical Incision: A Retrospective Study Authors: Xiaoxing Li, Lishan Huang, Yufei Li, Siyu Li, Xin Li, Xuechao Yang Published: 2026 ABSTRACT BackgroundDynamic navigation (DN) and static navigation (SN) in endodontic microsurgery (EMS) have been shown to improve the precision of osteotomy and root-end resection (RER). However, the impact of minimal incisions on DN- and SN-guided EMS remains unclear. This study assessed the accuracy of DN and SN in EMS performed via a single vertical incision at the apex (SVIA), and identified technique-specific accuracy predictors.MethodsA retrospective analysis included 60 teeth treated with SVIA EMS, comprising 30 osteotomies and RERs each in the SN and DN groups. Preoperative variables – sex, age, tooth position, arch type, drilling depth, lesion size, and buccal bone integrity – were recorded. Postoperative cone-beam computed tomography (CBCT) scans were superimposed with preoperative plans to evaluate accuracy. Group comparisons used independent t-tests or Mann–Whitney U tests (α = 0.01); general linear model examined the interaction effects between the technique and preoperative variables on accuracy; multivariate regression (α = 0.017) identified independent predictors within each group.ResultsNo significant differences in accuracy parameters were observed between DN and SN (P > .01). Interaction effects were nonsignificant (P > .05). Intragroup analysis indicated that for DN, a lesion diameter > 9 mm predicted increased resection angle deviation (RAD) (B = 3.13, P < .017), while for SN, posterior tooth location predicted reduced RAD (B = −5.80, P < .017).ConclusionsDN and SN exhibited comparable overall accuracy in SVIA EMS. However, DN was less accurate in cases with large lesions, whereas SN achieved higher accuracy in posterior regions. These preliminary findings support the evidence-based selection of guidance techniques according to anatomic specifics in SVIA EMS. Introduction Endodontic microsurgery (EMS) presents significant technical challenges in achieving precise osteotomy and root-end resection (RER).1 While the traditional freehand technique relies heavily on the surgeon’s experience and mental interpretation of cone-beam computed tomography (CBCT) images, computer-aided static (SN) and dynamic navigation (DN) systems have been introduced to enhance accuracy. SN utilizes a three-dimensionally (3D) printed surgical guide with a predefined trajectory, whereas DN employs real-time optical tracking for instrument guidance. Both modalities facilitate controlled execution along a virtual surgical plan, improving predictability in oral surgical procedures.2, 3, 4, 5, 6, 7, 8 Effective soft tissue management in EMS necessitates a balance between ensuring sufficient surgical access and achieving optimal wound closure, which is a prerequisite for precise EMS procedures.9 The single vertical incision at the apex (SVIA) with lateral flap reflection, a minimally invasive technique, significantly complicates freehand operations, resulting in unavoidable deviations in apical localization and RER, thereby adversely impacting surgical precision and long-term outcomes. However, the integration of computer-aided EMS with SVIA represents an innovative exploratory practice that merges suboptimal incision design with precision navigation technology. Recent case reports have highlighted the successful implementation of SN- and DN-guided EMS through SVIA, achieving precise osteotomy and RER.10, 11, 12, 13 Despite these promising findings, clinical evidence remains sparse. To date, only one retrospective series has assessed the accuracy of SN-guided RER via SVIA, with its conclusions limited by a small sample size and restricted outcome parameters.12 Crucially, no comparative clinical studies have investigated potential differences between SN and DN when employed with the SVIA technique, nor have any studies specifically evaluated the impact of anatomical factors on the accuracy of SVIA EMS. Therefore, this clinical study aimed to quantitatively evaluate and compare the accuracy of DN and SN for osteotomy and RER in EMS performed via SVIA. A secondary objective was to assess the influence of various clinical and anatomic factors – including tooth position, arch type, buccal bone integrity, lesion size, and drilling depth – on the achieved accuracy. Methods Study design This retrospective study was conducted at the Department of Endodontics, School and Hospital of Stomatology, Guangzhou Medical University (Guangzhou, China). Ethical approval was obtained from the Ethics Committee of Guangzhou Medical University (KY2019065), and the study was registered with the Chinese Clinical Trial Registry (ChiCTR2300074205). Informed consent was secured from all participants before their surgical intervention and enrolment in the study. The study adhered to the STROBE guidelines and complied with the World Medical Association’s Declaration of Helsinki. Sample size calculation A power analysis was performed using G*Power (version 3.1.9.7, Heinrich-Heine-Universität Düsseldorf, Germany) based on existing in vivo data for SN- and DN-guided EMS.14,15 The reported deviation in the resection angle was 9.711 ± 3.593° for SN and 4.9° for DN, yielding a mean difference of 4.811°. Given that the standard deviation for DN was unreported and clinical variability often exceeds that of published data, a conservative effect size of d = 0.9 was employed. With an effect size of 0.90, an α level of 0.05, and a β level of 0.10 (90% power), a minimum of 27 samples per group was determined to be necessary. To ensure statistical robustness, the final enrolment was increased to 30 participants per group. Consequently, the study included 30 roots in the SN group and 30 roots in the DN group. Patient selection Between January 2023 and December 2025, a retrospective identification of consecutive patients aged 18 to 60 years was conducted. These patients were diagnosed with previously treated teeth with symptomatic apical periodontitis and were indicated for guided EMS employing the SVIA approach. The inclusion criteria were as follows:1.Surgery for a single tooth (molar with only one affected root) without severe apical curvature (< 10°) according to the method of Schneider162.A tooth-supported template with a metal sleeve for SN-EMS3.Osteotomy and RER using a 4-mm conventional trephine 1.Surgery for a single tooth (molar with only one affected root) without severe apical curvature (< 10°) according to the method of Schneider16 Surgery for a single tooth (molar with only one affected root) without severe apical curvature (< 10°) according to the method of Schneider16 2.A tooth-supported template with a metal sleeve for SN-EMS A tooth-supported template with a metal sleeve for SN-EMS 3.Osteotomy and RER using a 4-mm conventional trephine Osteotomy and RER using a 4-mm conventional trephine The exclusion criteria comprised:1.Teeth with a prior history of endodontic surgery2.Teeth exhibiting pre-existing root fractures, perforations, or resorption 1.Teeth with a prior history of endodontic surgery Teeth with a prior history of endodontic surgery 2.Teeth exhibiting pre-existing root fractures, perforations, or resorption Teeth exhibiting pre-existing root fractures, perforations, or resorption A total of 62 consecutive patients who underwent SVIA EMS, utilizing either the DN (n = 30) or SN (n = 32) techniques, were enrolled following the application of inclusion and exclusion criteria. The analysis was conducted on the first 30 consecutive eligible cases from each technique group. Data collection Preoperative data were systematically gathered from the enrolled participants, including variables such as sex, age, tooth position (classified as anterior or posterior), arch type (maxillary or mandibular), condition of the apical buccal bone plate (categorized as intact or perforated), lesion size, and drilling depth. The dimension of the periapical lesion was assessed by measuring the maximum diameter of the radiolucent area situated in the apical or lateral region of the root using preoperative CBCT images. For each case, the CBCT dataset was reconstructed in three orthogonal planes (axial, sagittal, and coronal). The maximum lesion diameter was identified as the largest diameter in any of the three dimensions. Drilling depth was defined as the distance from the buccal cortical plate to the end of the planned trajectory. EMS procedures All surgical interventions were conducted by an endodontic specialist possessing a decade of experience in freehand EMS and 6 years of proficiency with SN and DN techniques. Before clinical application, 20 osteotomies and RERs were executed on in vitro models for each technique. SN-guided EMS protocol All procedures guided by SN utilized tooth-supported surgical templates with metal sleeves, as previously documented.12 In summary, preoperative CBCT scans were acquired using an iCAT 17-19 unit (Imaging Sciences International, Hatfield, PA, USA) with settings of 90 kVp, 5.0 mA, a field of view measuring 40 × 40 mm, and a voxel size of 0.09 mm. The CBCT data, formatted in Digital Imaging and Communications in Medicine (DICOM), and intraoral optical impressions obtained via a 3Shape scanner (Denmark), formatted in standard tessellation language (STL), were coregistered using 3Shape Implant Studio (D1000, Whip Mix Corp, Louisville, KY). A sleeve surgical template was virtually designed for a nonbeveled osteotomy and a 3-mm RER, followed by the fabrication of a resin template using a 3D printer (3510SD, 3D System Corporation, Rock Hill, SC, USA). Prior to surgery, its stability and adaptation were verified. An appropriate 8mm vertical incision was made using 15C blades (Hu-Friedy, Chicago, IL), extending from the vestibule to the attached gingiva, directly over the root in question, and aligned with the root’s long axis. After carefully elevating a full-thickness flap with retraction by the introducer port of the template, a template-guided osteotomy and RER were conducted using a 4-mm diameter trephine. The resected surface was conditioned with methylene blue and examined under a microscope (OPMI Proergo, Carl Zeiss, Göttingen, Germany). Root-end preparation was done ultrasonically, followed by retrofilling with iRoot BP Plus (Innovative Bioceramics, Vancouver, Canada). After cleaning the bone crypt and placing concentrated growth factor (CGF), the flap was repositioned and sutured with 6-0 monofilament sutures. Sutures were removed 1 week after surgery. DN-guided EMS protocol The DN procedures were conducted utilizing the SVIA approach in accordance with the established protocol.11 Preoperative CBCT images were obtained using exposure parameters identical to those of the SN-guided protocol. This process of marker-based registration for DN-EMS incorporated a silicone-based registration positioner embedded with radiopaque fiducial markers, which were secured to the contralateral molar regions. The CBCT data, formatted in DICOM, were imported into the digital dental navigation system (IRIS-100, Eped Inc., Taiwan, China) for the purpose of surgical path planning. The virtual surgical trajectory was designed to achieve a 3-mm resection of the root apex perpendicular to the root’s long axis, and this plan was subsequently entered into the DN interface. Following the calibration of the handpiece-reference device and the virtual-to-physical registration based on markers or feature points, the SVIA procedure was executed. A customized retractor was used to retract the flap, allowing for a real-time DN-guided osteotomy and RER with a 4-mm trephine. The resected surface was stained with methylene blue and examined microscopically before retrograde preparation was performed using an ultrasonic tip, followed by filling with iRoot BP Plus. After comprehensive debridement of the bone crypt and implantation of CGF, the flap margins were adapted and sutured. Sutures were removed at the 1-week postoperative interval. Postoperative CBCT examination Follow-up appointments were arranged at 1, 3, and 6 months for clinical evaluations and periapical radiographs, with CBCT conducted at the 1-year follow-up in accordance with the guidelines on CBCT usage in endodontics established by the European Society of Endodontology (ESE) and the ‘As Low As Reasonably Achievable’ (ALARA) principle. The CBCT scans were performed using the same equipment and settings as those utilized for the preoperative CBCT scans. To quantify accuracy, CBCT images from the follow-up visits were retrieved from the clinical database. Accuracy measurement All accuracy measurements were performed by an independent calibrated examiner who was blinded to the group allocation. To minimize random error and assess intraexaminer reliability, each parameter was measured twice with a 1-month interval, and the mean values were used for statistical analysis. Postoperative CBCT data (DICOM format) were imported into Mimics Research 21.0 software (Materialise NV, Leuven, Belgium) to generate a 3D model. A 4-mm virtual cylinder representing the actual surgical path was configured based on the border and contour of the resected root-end. The preoperative 3D model, including the virtual cylinder from the preoperative plan, was superimposed onto the postoperative 3D model using voxel-based registration anchored on stable anatomical structures. Following superimposition, the STL files were exported to Materialise 3-matic 11.0 (Materialise NV, Leuven, Belgium) for quantitative analysis, as described in a previous study.17 Five parameters were measured (Figure 1):Platform deviation (PD, mm): 3D distance between the planned and actual surgical trajectories at the labial or buccal cortical bone surface or root surface when the buccal bone plate was perforated.End deviation (ED, mm): 3D distance between the planned and actual surgical trajectories at the lingual termination point of the root.Angular deflection (AD, °): Angle between the central axes of the planned and actual surgical trajectories.Resection length deviation (RLD, mm): Difference between the planned resection length of 3 mm and the actual resection length along the root axis.Resection angle deviation(RAD, °): Angular discrepancy between the planned 90° bevel at the resection level along the long axis of the root and the actual angle of resection.Fig. 1The diagram of the accuracy parameters. A, Parameters of root-end resection. B, Parameters of osteotomy. A, the intersection of the root long axis and the planned trajectory; AD, angular deflection; AT, actual trajectory; B, the intersection of the root long axis and the actual trajectory; C, the planned entry point at the labial or buccal cortical bone surface; D, the actual entry point at the labial or buccal cortical bone surface; E, the lingual termination point of the actual surgical trajectory; ED, end deviation; F, the lingual termination point of the planned surgical trajectory; PD, platform deviation; PT, planned trajectory; RAD, resection angle deviation; RLD, resection length deviation.Fig 1 dummy alt text Platform deviation (PD, mm): 3D distance between the planned and actual surgical trajectories at the labial or buccal cortical bone surface or root surface when the buccal bone plate was perforated. Platform deviation (PD, mm): 3D distance between the planned and actual surgical trajectories at the labial or buccal cortical bone surface or root surface when the buccal bone plate was perforated. End deviation (ED, mm): 3D distance between the planned and actual surgical trajectories at the lingual termination point of the root. End deviation (ED, mm): 3D distance between the planned and actual surgical trajectories at the lingual termination point of the root. Angular deflection (AD, °): Angle between the central axes of the planned and actual surgical trajectories. Angular deflection (AD, °): Angle between the central axes of the planned and actual surgical trajectories. Resection length deviation (RLD, mm): Difference between the planned resection length of 3 mm and the actual resection length along the root axis. Resection length deviation (RLD, mm): Difference between the planned resection length of 3 mm and the actual resection length along the root axis. Resection angle deviation(RAD, °): Angular discrepancy between the planned 90° bevel at the resection level along the long axis of the root and the actual angle of resection. Resection angle deviation(RAD, °): Angular discrepancy between the planned 90° bevel at the resection level along the long axis of the root and the actual angle of resection. The diagram of the accuracy parameters. A, Parameters of root-end resection. B, Parameters of osteotomy. A, the intersection of the root long axis and the planned trajectory; AD, angular deflection; AT, actual trajectory; B, the intersection of the root long axis and the actual trajectory; C, the planned entry point at the labial or buccal cortical bone surface; D, the actual entry point at the labial or buccal cortical bone surface; E, the lingual termination point of the actual surgical trajectory; ED, end deviation; F, the lingual termination point of the planned surgical trajectory; PD, platform deviation; PT, planned trajectory; RAD, resection angle deviation; RLD, resection length deviation. Statistical analysis Statistical analyses were conducted using SPSS version 26.0 (IBM Corp, Armonk, NY, USA). The Intraclass Correlation Coefficient (ICC) analysis was employed to determine intraexaminer agreement. The Shapiro-Wilk test and the test for homogeneity of variance were applied to evaluate the normality of data distribution. For continuous variables with a normal distribution, data were reported as mean ± standard deviation (SD), while variables not conforming to a normal distribution were represented as median with interquartile range (IQR). In the primary analysis, comparisons between the DN and SN groups across the five accuracy parameters were conducted using the independent t-test for data exhibiting normal distribution, and the Mann-Whitney U test for data not conforming to a normal distribution. To control the family-wise error rate in these primary intergroup comparisons, a Bonferroni-corrected significance level of α = 0.01 was employed. Effect sizes were determined using Cohen’s d for parametric tests and Rank-Biserial correlation coefficients for nonparametric tests. A general linear model was used to evaluate the associations between the type of technique and preoperative variables. To explore potential effect modification, interaction terms between technique type and each preoperative variable were tested (using the Wald test) within the general linear model framework. A P value of < .05 for the interaction term was considered statistically significant. For the exploratory analysis, multivariate linear regression was applied to identify potential influencing factors within each group. Each accuracy parameter was evaluated concurrently with three preoperative variables. To address multiple comparisons within each parameter, a Bonferroni-corrected threshold of α = 0.017 (0.05/3) was applied. Results A total of 60 roots were included in the final analysis, with an equal allocation of 30 roots to each group. Figures 2 and 3 illustrate the treatment procedures associated with each technique. The baseline characteristics of the patients and teeth were comparable between the DN and SN groups. As presented in Table 1, there were no statistically significant differences between the groups concerning any preoperative variables, such as sex, age, tooth position, arch type, drilling depth, lesion size, and the status of the buccal bone plate (all P > .05). Lesion size was classified into large lesions (> 9 mm; n = 34) and small lesions (≤ 9 mm; n = 26) based on the mean length value. Drilling depth was categorized into > 6 mm (n = 25) and ≤ 6 mm (n = 35) using the median depth value as a reference.Fig. 2Surgical procedures using the DN system of tooth 8. A, Preoperative intraoral photography. B, Preoperative periapical radiography. C and D, Preoperative CBCT images. E, Presurgical path planning. F, Calibration. G, Registration. H, Single vertical incision at the apex with lateral retraction. I, Trephine bur placement. J, Real-time guidance of dynamic navigation. K, Osteotomy and root-end resection. L, Methylene blue staining. M, Retrograde filling. N, Suture. O, Immediate postoperative radiography. P and Q, 12-mo follow-up. R, 27-mo follow-up.Fig 2 dummy alt textFig. 3Surgical procedures using the SN technique of tooth 11. A, Preoperative intraoral photography. B, Preoperative periapical radiography. C and D, Preoperative CBCT images. E and F, 3D-printed template design. G, 3D-printed template check. H, Single vertical incision at the apex. I, 3D-printed template position and trephine bur placement. J, Osteotomy and root-end resection. K, Methylene blue staining. L, Retrograde filling. M, Suture. N, Immediate postoperative radiography. O and P, 14-mo follow-up.Fig 3 dummy alt textTable 1Baseline characteristics of the dynamic and static navigation groups.Table 1 dummy alt textSNDNVariablesn%n%P valueSex.084 Male1240.0516.7 Female1860.02583.3Age.505 Median (IQR)27 (8)29 (15.25)Tooth position.292 Anterior2066.71653.3 Posterior1033.31446.7Arch type1.000 Maxillary2273.32273.3 Mandibular826.7826.7Drilling depth.089 Median (IQR)5.15 (1.6)6.15 (2.98)Lesion diameter.297 > 9 mm1550.01963.3 ≤ 9 mm1550.01136.7Status of buccal bone plate.532 Intact2273.32583.3 Perforation826.7516.7DN, dynamic navigation; IQR, interquartile range; SN, static navigation. Surgical procedures using the DN system of tooth 8. A, Preoperative intraoral photography. B, Preoperative periapical radiography. C and D, Preoperative CBCT images. E, Presurgical path planning. F, Calibration. G, Registration. H, Single vertical incision at the apex with lateral retraction. I, Trephine bur placement. J, Real-time guidance of dynamic navigation. K, Osteotomy and root-end resection. L, Methylene blue staining. M, Retrograde filling. N, Suture. O, Immediate postoperative radiography. P and Q, 12-mo follow-up. R, 27-mo follow-up. Surgical procedures using the SN technique of tooth 11. A, Preoperative intraoral photography. B, Preoperative periapical radiography. C and D, Preoperative CBCT images. E and F, 3D-printed template design. G, 3D-printed template check. H, Single vertical incision at the apex. I, 3D-printed template position and trephine bur placement. J, Osteotomy and root-end resection. K, Methylene blue staining. L, Retrograde filling. M, Suture. N, Immediate postoperative radiography. O and P, 14-mo follow-up. Table 1Baseline characteristics of the dynamic and static navigation groups.Table 1 dummy alt textSNDNVariablesn%n%P valueSex.084 Male1240.0516.7 Female1860.02583.3Age.505 Median (IQR)27 (8)29 (15.25)Tooth position.292 Anterior2066.71653.3 Posterior1033.31446.7Arch type1.000 Maxillary2273.32273.3 Mandibular826.7826.7Drilling depth.089 Median (IQR)5.15 (1.6)6.15 (2.98)Lesion diameter.297 > 9 mm1550.01963.3 ≤ 9 mm1550.01136.7Status of buccal bone plate.532 Intact2273.32583.3 Perforation826.7516.7DN, dynamic navigation; IQR, interquartile range; SN, static navigation. Baseline characteristics of the dynamic and static navigation groups. DN, dynamic navigation; IQR, interquartile range; SN, static navigation. The ICCs for the DN group were 0.877, 0.860, 0.866, 0.909, and 0.901 for PD, ED, AD, RLD, and RAD, respectively. Meanwhile, the SN group exhibited ICCs of 0.827, 0.839, 0.918, 0.854, and 0.868, indicating a high degree of repeatability. Detailed accuracy metrics for both techniques are presented in Table 2. No statistically significant differences were detected between the DN and SN groups across any of the five accuracy parameters (all P > .01). Posthoc effect size analysis indicated trivial effect sizes (all Cohen’s d/Rank-Biserial r < 0.10) for PD, ED, and RLD, while small-to-moderate effect sizes were identified for AD (r = 0.231) and RAD (d = 0.314) (Table 5). Furthermore, analysis using a general linear model showed no interaction effects between the technique and subgroup variables (tooth position, arch type, drilling depth, lesion size, and the status of the buccal bone plate) on accuracy parameters (all P > .05).Table 2Comparison of accuracy parameters between SN and DN groups.Table 2 dummy alt textParametersnSNDNP valuePD/mm600.62 ± 0.310.71 ± 0.36.290ED/mm600.67 (0.34)0.69 (0.42).819AD/°607.11 (4.65)5.97 (3.97).204RLD/mm600.61 (0.38)0.56 (0.42).882RAD/°606.53 ± 3.985.17 ± 2.88.136AD, angular deflection; DN, dynamic navigation; ED, end deviation; PD, platform deviation; RAD, resection angle deviation; RLD, resection length deviation; SN, static navigation.Descriptive statistics are presented as mean ± standard deviation when the data conformed to a normal distribution;otherwise, the median (interquartile range) is used.A Bonferroni correction determined data to be significant at P < .01. Table 2Comparison of accuracy parameters between SN and DN groups.Table 2 dummy alt textParametersnSNDNP valuePD/mm600.62 ± 0.310.71 ± 0.36.290ED/mm600.67 (0.34)0.69 (0.42).819AD/°607.11 (4.65)5.97 (3.97).204RLD/mm600.61 (0.38)0.56 (0.42).882RAD/°606.53 ± 3.985.17 ± 2.88.136AD, angular deflection; DN, dynamic navigation; ED, end deviation; PD, platform deviation; RAD, resection angle deviation; RLD, resection length deviation; SN, static navigation.Descriptive statistics are presented as mean ± standard deviation when the data conformed to a normal distribution;otherwise, the median (interquartile range) is used.A Bonferroni correction determined data to be significant at P < .01. Comparison of accuracy parameters between SN and DN groups. AD, angular deflection; DN, dynamic navigation; ED, end deviation; PD, platform deviation; RAD, resection angle deviation; RLD, resection length deviation; SN, static navigation. Descriptive statistics are presented as mean ± standard deviation when the data conformed to a normal distribution; otherwise, the median (interquartile range) is used. A Bonferroni correction determined data to be significant at P < .01. In the DN group, as presented in Table 3, a lesion diameter greater than 9 mm was determined to be a significant independent predictor of increased RAD, with a coefficient of B = 3.13 (95% CI: 1.14-5.12, P < .017). Conversely, in the SN group, as shown in Table 4, the presence of posterior teeth was identified as a significant predictor of reduced RAD, with a coefficient of B = −5.80 (95% CI: −9.63 to −1.96, P < .017).Table 3Multivariate regression analysis of factors associated with accuracy in dynamic navigation group.Table 3 dummy alt textFactorsPD (mm)ED (mm)AD (°)RLD (mm)RAD (°)Tooth positionB−0.22−0.13−1.16−0.19−1.56 Posterior vs anterior95% CI−0.50 to 0.07−0.44 to 0.18−3.28 to 0.97−0.46 to 0.08−3.53 to 0.40P value.127.385.271.158.114Lesion diameterB0.130.262.420.153.13 >9 mm vs ≤9 mm95% CI−0.16 to 0.42−0.05 to 0.580.26-4.57−0.13 to 0.421.14-5.12P value.353.096.029.285.003*Status of buccal bone plateB−0.29−0.35−0.15−0.33−0.01 Perforation vs intact95% CI−0.66 to 0.08−0.76 to 0.06−2.95 to 2.65−0.69 to 0.03−2.60 to 2.57P value.121.088.911.067.992AD, angular deflection; B, beta coefficient; CI, confidence interval; ED, end deviation; PD, platform deviation; RAD, resection angle deviation; RLD, resection length deviation.The model was adjusted for age and sex.A Bonferroni correction determined data to be significant at P < .017.⁎indicates statistical significance at P < .017.Table 4Multivariate regression analysis of factors associated with accuracy in static navigation group.Table 4 dummy alt textFactorsPD (mm)ED (mm)AD (°)RLD (mm)RAD (°)Tooth positionB−0.18−0.31−4.47−0.30−5.80 Posterior vs anterior95% CI−0.53 to 0.17−0.74 to 0.13−8.96 to 0.03−0.60 to 0.01−9.63 to −1.96P value.297.163.051.054.005*Arch typeB0.170.285.080.203.87 Mandibular vs Maxillary95% CI−0.19 to 0.53−0.18 to 0.730.41-9.76−0.12 to 0.51−0.20 to 7.87P value.345.217.034.208.057Drilling depthB−0.02−0.021.580.0033.23 > 6mm vs ≤ 6 mm95% CI−0.38 to 0.34−0.48 to 0.43−3.06 to 6.22−0.31 to 0.32−0.74 to 7.19P value.926.913.489.985.106AD, angular deflection; B, beta coefficient; CI, confidence interval; ED, end deviation; PD, platform deviation; RAD, resection angle deviation; RLD, resection length deviation.The model was adjusted for age and sex.A Bonferroni correction determined data to be significant at P < .017.⁎indicates statistical significance at P < .017.Table 5Effect size analysis.Table 5 dummy alt textParametersCohen’s dRank-Biserial rPD0.051-ED-0.034AD-0.231RLD-0.022RAD0.314- Table 3Multivariate regression analysis of factors associated with accuracy in dynamic navigation group.Table 3 dummy alt textFactorsPD (mm)ED (mm)AD (°)RLD (mm)RAD (°)Tooth positionB−0.22−0.13−1.16−0.19−1.56 Posterior vs anterior95% CI−0.50 to 0.07−0.44 to 0.18−3.28 to 0.97−0.46 to 0.08−3.53 to 0.40P value.127.385.271.158.114Lesion diameterB0.130.262.420.153.13 >9 mm vs ≤9 mm95% CI−0.16 to 0.42−0.05 to 0.580.26-4.57−0.13 to 0.421.14-5.12P value.353.096.029.285.003*Status of buccal bone plateB−0.29−0.35−0.15−0.33−0.01 Perforation vs intact95% CI−0.66 to 0.08−0.76 to 0.06−2.95 to 2.65−0.69 to 0.03−2.60 to 2.57P value.121.088.911.067.992AD, angular deflection; B, beta coefficient; CI, confidence interval; ED, end deviation; PD, platform deviation; RAD, resection angle deviation; RLD, resection length deviation.The model was adjusted for age and sex.A Bonferroni correction determined data to be significant at P < .017.⁎indicates statistical significance at P < .017. Multivariate regression analysis of factors associated with accuracy in dynamic navigation group. AD, angular deflection; B, beta coefficient; CI, confidence interval; ED, end deviation; PD, platform deviation; RAD, resection angle deviation; RLD, resection length deviation. The model was adjusted for age and sex. A Bonferroni correction determined data to be significant at P < .017. indicates statistical significance at P < .017. Table 4Multivariate regression analysis of factors associated with accuracy in static navigation group.Table 4 dummy alt textFactorsPD (mm)ED (mm)AD (°)RLD (mm)RAD (°)Tooth positionB−0.18−0.31−4.47−0.30−5.80 Posterior vs anterior95% CI−0.53 to 0.17−0.74 to 0.13−8.96 to 0.03−0.60 to 0.01−9.63 to −1.96P value.297.163.051.054.005*Arch typeB0.170.285.080.203.87 Mandibular vs Maxillary95% CI−0.19 to 0.53−0.18 to 0.730.41-9.76−0.12 to 0.51−0.20 to 7.87P value.345.217.034.208.057Drilling depthB−0.02−0.021.580.0033.23 > 6mm vs ≤ 6 mm95% CI−0.38 to 0.34−0.48 to 0.43−3.06 to 6.22−0.31 to 0.32−0.74 to 7.19P value.926.913.489.985.106AD, angular deflection; B, beta coefficient; CI, confidence interval; ED, end deviation; PD, platform deviation; RAD, resection angle deviation; RLD, resection length deviation.The model was adjusted for age and sex.A Bonferroni correction determined data to be significant at P < .017.⁎indicates statistical significance at P < .017. Multivariate regression analysis of factors associated with accuracy in static navigation group. AD, angular deflection; B, beta coefficient; CI, confidence interval; ED, end deviation; PD, platform deviation; RAD, resection angle deviation; RLD, resection length deviation. The model was adjusted for age and sex. A Bonferroni correction determined data to be significant at P < .017. indicates statistical significance at P < .017. Table 5Effect size analysis.Table 5 dummy alt textParametersCohen’s dRank-Biserial rPD0.051-ED-0.034AD-0.231RLD-0.022RAD0.314- Effect size analysis. Discussion This study evaluated the impact of SVIA on the accuracy of DN and SN in EMS. The primary finding is that, within the SVIA framework, both DN and SN provided statistically equivalent overall accuracy for osteotomy and RER, affirming their viability as precise, minimally invasive options. However, intragroup analyses revealed that the constrained surgical access differentially modulated the performance of each technology, leading to distinct, technique-specific predictors of accuracy. Comparative accuracy in the context of SVIA The use of CBCT in EMS for assessing postsurgical healing outcomes is both accepted and recommended due to its sensitivity and specificity. According to ESE guidelines on CBCT usage in endodontics and adhering to the ALARA principle, CBCT is recommended for use at a 1-year follow-up interval. In consideration of radiation exposure, this study assessed five accuracy parameters using presurgical planning and postoperative CBCT at the 1-year follow-up. Currently, only four in vivo studies investigate the accuracy of computer-aided EMS, with three employing postoperative 1-year follow-up CBCT and one utilizing postoperative 1-week CBCT.12,14,15,18 In our study, to ensure reliable superimposition, osteotomy and RER were performed using a trephine bur, reshaping the resected apex into a concave configuration. The direction and angle of the actual surgical trajectory, following 1 year of bone healing, were reconstructed by mapping the postoperative virtual cylinder based on the border and contour of the resected root-end, thereby facilitating sufficient reliable superimposition. Prior simulation studies consistently report superior accuracy for both SN and DN over freehand technique, with no significant difference between the two guided methods, although SN is often associated with shorter operative time.17,19, 20, 21, 22 Our clinical results, utilizing comprehensive deviation metrics (entry/apex/angle deviation for osteotomy; resection length/angle deviation for RER), align with these findings. However, the moderate effect sizes observed for AD and RAD in this study indicate the potential for a clinically meaningful difference, thereby justifying further investigation with a larger sample size. When compared to existing clinical reports on flapped EMS, our observed accuracy values for both DN and SN fall within a clinically acceptable range, albeit with some variation.14,15,18 These differences are likely attributable to disparities in study design, case selection, surgical protocols, and navigation systems, beyond the specific effect of SVIA. Notably, only one prior retrospective series has evaluated SN-RER accuracy specifically via SVIA, reporting higher angular deviation.12 The improved accuracy in our SN cohort may be partially explained by our single-operator, experienced-surgeon design. Collectively, the evidence indicates that both guided techniques can achieve high precision in SVIA EMS. Technique-specific determinants of accuracy with SVIA The precision mechanisms of SN and DN are fundamentally distinct, and the SVIA approach uniquely influences these mechanisms. For SN, accuracy is critically dependent on the stable, passive fit of the surgical guide.23 Contrary to the conventional concern that posterior anatomy may challenge guide placement, our data indicated that posterior teeth were associated with reduced RAD in the SN-SVIA group. This counterintuitive finding underscores the profound impact of minimally invasive access. The denser, flatter buccal bone in the posterior region likely provides a more stable foundation for guide seating compared to the thinner, more curved anterior anatomy. Furthermore, the mucosal characteristics in an SVIA design may play a role; the greater dissection involving the more mobile alveolar mucosa in the anterior region could compromise guide stability more than in the posterior region. Additionally, the restricted visual field in SVIA limits the surgeon’s ability to visually detect and correct minor guide instability or tissue springback during trephine alignment, potentially magnifying errors. For DN, accuracy is primarily governed by registration precision and the surgeon’s ability to correlate the navigation display with the operative field.24,25 Since DN relies on virtual visualization rather than direct sight of the root apex, restricted visual access inherent to SVIA – particularly in the posterior region – does not inherently impede its system performance. This is supported by our findings and others that tooth position did not significantly affect DN accuracy.19,24,26 DN’s strength lies in allowing real-time positional feedback and trajectory adjustment, making it advantageous for anatomically complex sites. However, this approach demands proficient hand-eye coordination to transition between the virtual plan and the surgical site.27 While DN can improve novice performance, it may not fully bridge the gap with expert freehand precision, suggesting that surgeon experience remains relevant.28 In our study, performed by an experienced surgeon, DN accuracy was consistent across regions. A key predictor specific to the DN group was lesion size. A lesion diameter > 9 mm was associated with increased RAD. We hypothesize that extensive osteolytic defects may reduce the stabilizing effect of surrounding bone on the trephine during DN-guided resection, potentially exacerbating minor positional tremor that a surgeon might otherwise compensate for under direct vision or with a physically constrained SN guide. To identify technique-specific determinants of accuracy in SVIA through multivariate logistic regression analysis, each accuracy parameter was evaluated concurrently with three preoperative variables: tooth position, arch type, and drilling depth, which were included based on a prior study.19 For the DN group, buccal bone integrity and lesion size were also considered, as previous research has suggested that the trephine used during the initial drilling process poses a potential risk of bur slippage, particularly at the entry point, potentially affecting DN accuracy.29 Clinical implications and technical considerations The integration of SVIA with guided EMS epitomizes a minimally invasive philosophy but introduces specific challenges, including limited maneuvering space and soft tissue retraction difficulties. Several technical refinements can optimize outcomes: the use of a metal sleeve or tooth-and-bone-supported guide for enhanced SN stability; a customized 3D-printed retractor for improved SVIA access; and a ‘pilot’ or ‘stop’ trephine to prevent bur slippage and control resection depth.29, 30, 31, 32 Based on our findings, preliminary evidence suggests that DN may be preferentially considered for cases involving anatomically complex posterior teeth or where intraoperative path adjustment is anticipated, though caution is advised in the presence of very large periapical lesions (> 9 gt; 9 mm). SN appears to be a robust and efficient choice, particularly for teeth with favourable buccal bone topography (eg, posterior region or flat surface) that ensure guide stability, even with limited SVIA access. Limitations and future directions This study has limitations inherent to its retrospective, single-operator design, which may affect the generalizability of the results. The sample size, while robust for a clinical navigation study, may limit the detection of subtle between-group differences. Therefore, the exploratory regression findings require validation in larger, prospective studies. Furthermore, the evaluation focused on quantitative accuracy metrics;future prospective studies incorporating standardized healing outcomes and patient-reported measures are essential to fully assess the clinical efficacy of these guided SVIA protocols. Continued technical innovations, such as simplified CBCT-only guide fabrication and enhanced registration algorithms, hold promise for broadening the accessibility and reliability of both SN and DN in minimally invasive endodontic surgery.33 Conclusion Within the limitations of this retrospective study, both DN and SN demonstrated clinically acceptable accuracy for osteotomy and RER when performed via SVIA. This integration represents a reliable technological advancement for minimally invasive endodontic microsurgery. Based on these preliminary findings, the choice between DN and SN in this context should be a strategic one, guided by specific anatomic factors: SN may be advantageous in posterior sites with favourable bone topography to ensure guide stability, whereas DN offers particular utility in anatomically complex regions, with due consideration given to lesion size as a potential modifier of its precision. Author contributions Xiaoxing Li: Writing – Original Draft, Methodology, Data analysis. Lishan Huang: Methodology, Investigation,Validation, Data curation. Yufei Li: Methodology, Data curation. Siyu Li: Resources, Validation. Xin Li: Resources, Validation. Xuechao Yang: Conceptualization, Writing – review & editing, Supervision, Project administration, Funding acquisition. Conflict of interest None disclosed.