Role of periodontal endoscopy in clinical decision-making for teeth with suspected vertical root fractures Authors: Min Yang, Xuefei Sun, Gaofeng Fang, Yunpeng Xue, Qianqian Dong Published: 2025 ABSTRACT AbstractBackgroundThis study aimed to evaluate the role of periodontal endoscopy in clinical decision-making for teeth with suspected vertical root fracture (S-VRF) and to assess the diagnostic accuracy of this procedure with respect to vertical root fracture (VRF).MethodsA prospective cohort of 115 S-VRFs meeting the inclusion criteria was analysed. The teeth were independently evaluated by two endodontists and subsequently categorized into three groups based on the type of pre-endoscopic treatment: flap surgery (S group, n = 56), endodontic treatment (R group, n = 48), and extraction (E group, n = 11). Periodontal endoscopic examination was performed for all the teeth, and treatment plans were reassigned based on findings. The diagnosis of VRF was confirmed by intraoperative crack identification or clinical failure within a 1-year follow-up period. Three cases were excluded because they were lost to follow-up.ResultsCompared with cone-beam computed tomography(CBCT), periodontal endoscopy was associated with a reduction in unnecessary nonsurgical treatment by 39.58% (19/48), reduced the incidence of unnecessary surgical treatment by 64.28% (36/56), and an increase in diagnostic accuracy by 49.1%.The sensitivity of periodontal endoscopy for VRF detection was 84.3%.ConclusionsPeriodontal endoscopy may enhanced clinical decision-making and was associated with a reduced incidence of unnecessary nonsurgical and surgical interventions in patients with S-VRFs.Keywords: Vertical root fracture, Periodontal endoscopy, Diagnostic accuracy, Minimally invasive dentistry, Clinical decision-making Background Vertical root fracture (VRF) is defined as longitudinally oriented complete or incomplete fractures extending through the root structure at any level [1]. These fractures present significant diagnostic and therapeutic challenges in contemporary endodontics. Advanced VRF typically present with significant fracture displacement and extensive alveolar bone resorption at the fracture interface, resulting in a poor prognosis. Therefore, early detection of VRF is paramount for tooth preservation [2, 3]. Clinically, VRF manifest as sinus tracts, recurrent occlusal pain, narrow and deep (>6 mm) periodontal pockets, and other atypical symptoms [4–9], thus making diagnosis difficult [10]. Periapical radiographs may reveal abnormal widening of the root canal [11], but the efficacy of X-ray imaging for diagnosing VRF is limited by the imaging angle and overlapping tissues [12]. Cone-beam computed tomography (CBCT) images may reveal alveolar bone resorption along hypodense fracture lines, referred to as “halo” or “J” translucency [5, 6]. CBCT is more accurate than X-ray imaging for diagnosing VRF [13–17]; however, CBCT still has limitations. Smaller voxels and field-of-view sizes provide higher detection accuracy for VRF [18]. For small VRF, the accuracy of CBCT in vivo is only 29% [19]. Root canal filling materials or crown-square restorations can block projections and create artefacts in or blur out fracture lines [11]. Therefore, the diagnostic accuracy of CBCT for teeth with no root canal treatment is 2.399 times greater than that for teeth with root canal treatment [20]. Suspected vertical root fracture(S-VRF) has overlapping clinical manifestations with VRF, but no obvious cracks are observed on CBCT [21]. S-VRF may present as localized periodontal disease, combined periodontal‒pulpal lesions, early VRF, and transverse root fractures. The diagnosis of S-VRF is essential, as failing to diagnose VRFs in S-VRFs can lead to increased alveolar bone resorption. However, the overdiagnosis of non-VRF can lead to unnecessary tooth extractions. When assessing S-VRF, diagnostic protocols mandate CBCT evaluation following clinical examinations, and treatment decisions are contingent upon radiographic findings and periodontal parameters [21–24]. Definitive fracture visualization warrants tooth extraction [25]; teeth with compromised periodontal support (probing depth (PD) ≥ 6 mm with mobility) are scheduled for extraction; teeth with adequate bone support and inadequate root canal treatment (RCT) undergo RCT/root canal retreatment (re-RCT) and periodontal treatment, and teeth with persistent apical pathology despite adequate obturation are recommended for surgical intervention. The gold standard treatment for VRF is flap surgery [26]; however, this is an invasive surgical procedure [7]. Therefore, the current study aimed to identify a non-invasive and easy way to assess VRF. A periodontal endoscope is an optical instrument composed of a cold light source probe, fibre optic cables, image transmission systems, and screen display systems. By inserting the cold light source lens into the periodontal pocket, magnified images of the root surface can be viewed on a display to check for any obvious crack lines. The cold light source probe of the periodontal endoscope is designed to be compact, thus enabling a minimally invasive examination of the root surface. Zhang et al. [27] used a periodontal endoscope to diagnose VRF and achieved a correct index of 83.7%, which is higher than the 53.1% accuracy rate achieved with CBCT. The current study aimed to investigate the clinical utility of periodontal endoscopy for therapeutic decision-making and to assess its diagnostic accuracy for S-VRF. Methods This study was approved by the Institutional Review Board of Xi’an Jiaotong University Stomatological Hospital (No. KY-QX-20240004). Verbal and written informed consent was obtained from all the patients included in the study. From July 2022 to June 2023, patients with S-VRFs who were treated at the Department of Endodontics, Stomatology Hospital, Xi’an Jiaotong University, were enrolled if they met the following inclusion criteria: aged between 18 and 70 years; presented with at least one clinical symptom indicative of potential endodontic-periodontal pathology, specifically recurrent gingival swelling, a sinus tract, or occlusal pain; presented with an isolated periodontal pocket depth exceeding 6 mm, as measured with a UNC-15 probe (Hu-Friedy); and exhibited no evidence of definitive fracture lines on two-plane CBCT reconstructions (acquired using a NewTom VGi scanner at 90 kV and 8.9 mA). The exclusion criteria were as follows: the presence of systemic contraindications, including HbA1c levels > 7.5% or uncontrolled hypertension; a diagnosis of generalized periodontitis, characterized by ≥ 30% of sites exhibiting clinical attachment loss (CAL) ≥ 5 mm; existing root fractures or iatrogenic cracks identified clinically or radiographically; or current pregnancy or lactation. The sample size was calculated using diagnostic test assumptions (α = 0.05, β = 0.20, expected sensitivity = 85% [27, 28]) based on pilot data, requiring ≥ 100 evaluable cases. A total of 112 patients (45 males and 67 females; mean age, 44.6 ± 13.4 years) with 115 S-VRFs were included in the study. Diagnostic workflow One endodontist and one radiologist jointly assessed whether the teeth met the inclusion and exclusion criteria. Disagreements were resolved by consulting a senior surgeon. Two endodontists evaluated the S-VRFs, and the most appropriate treatment plan was selected based on the root canal filling status, restoration status, and periodontal status. The teeth were divided into three groups based on the treatment: flap surgery (S group), RCT/re-RCT endodontic treatment (R group), or extraction (E group). All teeth were subjected to periodontal endoscopy. The periodontal pockets of S-VRFs were infiltrated and anaesthetized with 4% articaine with 1:100,000 epinephrine; the periodontal endoscope was held with the right hand; the shield was inserted into the periodontal pocket; and the condition of the tooth at that site was visually assessed. Then, the endoscope shield from the periodontal pocket was removed, the manual scaler was held with the right hand, and root planing was performed to remove calculus, plaque, and diseased cementum. The root surface was stained with methylene blue and observed using the cold light source probe of a periodontal endoscope (P100, Bangwo, China). The ‘observation’-‘scaling’-‘staining’-‘reobservation’ steps were then repeated. All endoscopic examinations were performed by periodontists with more than 10 years of experience, and the results were interpreted jointly by one periodontist and one endodontist. The presence of continuous blue lines exceeding 2 mm in length on the periodontal endoscope screen indicates a vertical root fracture. Teeth with endoscopically observed cracks (Fig. 1d) and those in the E group were extracted or underwent root amputation when the periodontal condition of the roots of the maxillary molars was still favourable [28]. In the surgical group, teeth in which no cracks were observed endoscopically were explored via flap surgery. The root surfaces of these teeth were stained with methylene blue and observed under a microscope (OMS2380, Zumax, China). If cracks were observed, the tooth was considered to have a verified VRF, thus requiring extraction or root amputation. Teeth without cracks (Fig. 1c) were confirmed to be free of VRF, and surgery was performed. For teeth in the R group, the filling material in the root canal was removed, and methylene blue was used to stain the pulp chamber [21]. The upper one-third of the root canal or a root canal sizer was used for detection. If cracks were observed or if the sizer suggested an abnormality, the tooth was scheduled for extraction or root amputation. Teeth with cracks detected during root amputation were confirmed to have a VRF. Teeth with no obvious abnormality were confirmed to be free of VRF; a root canal was performed on these teeth, and they were followed up for 1 year. If the symptoms did not resolve during that time and a root fracture was detected at a later date, the teeth were considered to have VRF [26]. (a) Diagram of the use of a periodontal endoscope; (b) Intraoral use of periodontal endoscopes; (c) No cracks observed by endoscopy; (d) A crack observed by endoscopy Teeth were extracted by an experienced senior practitioner to minimize the possibility of intraoperative root fracture and to avoid excessive rotational forces. All the isolated teeth were examined microscopically for the presence or absence of cracks. The efficacy of periodontal endoscopy for VRF was evaluated by measuring the following parameters: the reduction in the rate of root canal therapy (endoscopically determined VRFs in the nonsurgical group/total nonsurgical group), the reduction in the rate of surgical treatment (endoscopically determined VRFs in the surgical group/total surgical group), sensitivity ([# patients with endoscopically determined VRFs/total # patients with VRFs] × 100%) and specificity ([# patients endoscopically determined to lack VRFs/total # patients without VRFs] × 100%). Inter- and intra-examiner agreement was assessed by calculating the kappa coefficient, and the relationships between multiple factors and VRF were assessed via logistic regression. Statistical analysis was performed using SPSS 22.0 software. CASE 1 A 64-year-old male presented with a 1-year history of persistent occlusal pain in the left mandibular posterior region. The patient reported root canal treatment (RCT) performed on tooth #19 four years prior, with no history of dental trauma or parafunctional habits. The medical history was noncontributory, with no systemic contraindications to dental procedures. Fillings and overhangs were observed on the distal and mesial adjacent surfaces of tooth #19, together with percussion tenderness and gingival redness and swelling. The proximal buccal periodontal pockets were probed to a depth of 16 mm. No sinus tracts were observed in the buccal alveolar mucosa, and the tooth was not mobile. On CBCT, there was no continuous hypodense line across the root structure, but endoscopy revealed obvious crack lines on the proximal buccal root surface. After extraction of tooth #19, proximal buccal root cracks were visible via microscopy (Fig. 2). (a) Sinus tract visible on the buccal side of tooth #19; (b) Proximal buccal exploration of a 16-mm periodontal pocket; (c–f) “J” type alveolar bone resorption in the proximal buccal root; (g) Endoscopic examination of the proximal buccal root revealed a crack; (h) A crack on the proximal buccal root Results The study population included 112 patients (45 males and 67 females; mean age,44.6 ± 13.4 years) with 115 S-VRFs. However, three S-VRFs were excluded from the statistical analysis because the patients did not complete the 1-year follow-up assessment. The 112 included S-VRFs (Table 1) were molars (n = 86, 76.8%), premolars (n = 25, 22.3%), and anterior teeth (n = 1, 0.9%). Prior RCT had been performed on 101 teeth (90.2%). Table 1Prevalence of VRF (with 95% CIs) by independent factorsVariablesn/N(%)P valueAge0.362 ≤3022/112(19.6%) 31~4022/112(19.6%) 41~5020/112(17.9%) 51~6020/112(17.9%) ≥6128/112(25.0%)Sex0.367 Female67/112(59.8%) Male45/112(40.2%)Tooth type0.168 Anterior teeth1/112(0.9%) Maxillary premolar15/112(13.4%) Maxillary molar15/112(13.4%) Mandibular premolar10/112(8.9%) Mandibular molar71/112(63.4%)ETT0.07 ETT*101/112(90.2%) NETT*11/112(9.8%)Restoration0.18 Unrestored15/112 (13.4%) Resin23/112(20.5%) Crown74/112(66.1%)PD(mm)- <8 mm0/112(0.0%) ≥8 mm112/112(100%)Mobility0.99 Normal79/112(70.5%) Ⅰ23/112(20.5%) Ⅱ8/112(7.2%) Ⅲ2/112(1.8%)Sinus tracts0.47 028/112(25.0%) 178/112(69.6%) 26/112(5.4%)Root canal filling0.077 Emptiness11/112(9.8%) Incomplete36/112(32.2%) Complete65/112(58.0%)*ETT:Endodontically treated teeth,;NETT:Non-endodontic treatment Prevalence of VRF (with 95% CIs) by independent factors *ETT:Endodontically treated teeth,;NETT:Non-endodontic treatment In the S group (n = 56), after periodontal endoscopic examination, one maxillary first molar underwent root resection due to a discovered root fracture in the buccal root, whereas the alveolar bone around the other roots was still intact. In the S-1 group (n = 20), 14 teeth showed no cracks and underwent apicoectomy. One tooth was extracted because of an extensive alveolar bone defect, one tooth underwent repair for a perforation, three teeth were extracted because of cracks discovered during surgery, and one tooth was extracted because of a crack discovered during the one-year postoperative observation period. The teeth in the R-1 group (n = 29) underwent endodontic treatment. Twenty of these teeth had no root cracks and had a good prognosis. Three teeth were extracted because of buccolingual cracks in the pulp chamber floor, one was extracted because of a crack in the mesiolingual canal, one was extracted because of an abnormal apex locator reading, one developed a root fracture four months after treatment, and three were lost to follow-up. In the E group (n = 11), no cracks were observed on the surface of five teeth, one had a root surface defect, and cracks were observed on the root surface of five teeth. In both the E-1 and E-2 groups, cracks were observed on the root surface of all the included teeth after extraction. An examination of the 112 S-VRFs revealed 70 VRFs (62.5%, 70/112) (Fig. 3). These VRFs were identified through the following modalities: periodontal endoscopy (n = 59, 84.3%), surgical flap reflection (n = 3, 4.3%), crown removal (n = 3, 4.3%), and RCT (n = 3, 4.3%). Two additional VRFs (n = 2, 2.9%) in the E group were diagnosed post-extraction. Overall, the identification of S-VRFs via periodontal endoscopy reduced the need for surgical treatment by 64.3% (36/56) and reduced the need for RCT/re-RCT by 42.2% (19/45). The clinical accuracy for diagnosing VRF improved by 49.1% (55/112) (Fig. 4). Enrolment and follow-up Changes in the treatment plan between CBCT and periodontal endoscopy examination for S-vRfs. *s group: lap surgery; R group: endodontic treatment; E group: extraction Interobserver agreement: CBCT alone showed moderate interobserver agreement (Cohen’s κ = 0.444), and CBCT combined with periodontal endoscopy showed a significantly higher level of interobserver agreement (Cohen’s κ = 0.842). The sensitivity of periodontal endoscopy for identifying VRF was 76.0% (19/25) (95% CI: 54.6% - 90.2%) for teeth in the ETT group, 92.5% (37/40) (95% CI: 79.6% - 98.4%) for teeth in the S group, and 84.3% (59/70) (95% CI: 75.0% - 91.1%) for teeth in the periodontal endoscopy group. The area under the curve (AUC) was 0.92 (Fig.5).The specificity of periodontal endoscopy for identifying VRF was 100%(59/59) (95% CI: 93.29%-100%).The Positive Likelihood Ratio (LR+) approached infinity(95% CI: 11.33-∞).The Negative Likelihood Ratio (LR-) was approximately 0.1571(95% CI, 0.0650-0.2572). In terms of predictive accuracy, the Positive Predictive Value (PPV) was 100% (95% CI: 93.97%-100%). The Negative Predictive Value (NPV) stood at approximately 79.25%(95% CI: 67.80%-90.49%). Receiver operating characteristic (ROC) curve for evaluating the diagnostic performance of periodontal endoscopy in detecting VRF Discussion The findings of this study indicated that preoperative periodontal endoscopic evaluation for S-VRFs had a sensitivity of 84.3% and an AUC of 0.92 for diagnosing VRF. LR+ approaches infinity, signifying that a positive test result virtually guarantees the presence of the target condition. LR- is approximately 0.1571, indicating a substantial 84.29% reduction in the odds of the condition when the test result is negative. In terms of predictive accuracy, PPV is 100%, confirming that every positive result is truly indicative of the disease. NPV stood at approximately 79.25%, meaning 79.25% of negative results can be confidently interpreted as excluding the condition. Compared with CBCT, preoperative periodontal endoscopy was associated with a reduced need for flap surgery by 64.28%, a reduced need for RCT by 42.22%, and resulted in a 49.1% increase in diagnostic accuracy. However, due to the small sample size of this experimental study, further research is needed to validate this viewpoint. Directly inserting the periodontal endoscope probe into the periodontal pocket to visualize the root surface is a non-invasive technique that eliminates the need for crown removal, root canal filling removal, or flap surgery, thus aligning with modern minimally invasive dental principles. When a fracture is confirmed via endoscopy, strategic tooth extraction can preserve more alveolar bone, thus providing a foundation for future restoration. Conversely, the absence of a detectable fracture allows for tooth preservation via RCT or surgery, which is consistent with contemporary biologically oriented treatment philosophies [29]. Currently, the main clinical methods for diagnosing VRF include radiographic examination, CBCT, RCT, flap surgery, and tooth extraction. The advantages of radiography include the low level of radiation and simplicity, but this approach is only applicable for fractures with significant fragment displacement and is susceptible to projection angle artefacts. CBCT is non-invasive and provides higher diagnostic accuracy than radiography; however, root-filling materials and coronal restorations often cause artefacts, thereby obscuring fracture visualization. Furthermore, the sensitivity of CBCT is lower for non-through-and-through fractures and cracks measuring <0.2 mm. RCT indicates high specificity for detecting through-and-through VRF but requires the removal of pulp tissue or fillings and involves root canal wall preparation, thereby compromising remaining tooth structure strength and increasing fracture risk. Flap surgery and tooth extraction remain the diagnostic gold standards for VRF, as these approaches offer high diagnostic efficacy. However, flap surgery is an invasive procedure that is typically used to confirm fracture presence rather than as a routine screening tool. Extraction is irreversible, contradicts the prevailing philosophy of preserving natural teeth, and involves multiple visits, causing patient inconvenience. Periodontal endoscopy not only enables visualization of VRFs but also aids in detecting root caries, horizontal root fractures, and root perforations. Its standardized procedure (scaling-staining-observation) offers strong clinical operability, and it can be performed chairside, thus reducing patient anxiety. Nevertheless, periodontal endoscopy has some limitations. The diagnostic efficacy of this approach for apical fractures, interproximal fractures, furcation perforations, and fractures associated with complex pockets is limited by the length and width of the endoscope sheath. Furthermore, owing to the resolution of the probe, early-stage cracks poorly stained by methylene blue may be missed. It is necessary for operators to have excellent spatial awareness, hand-eye coordination, and extensive periodontal experience. Additionally, the high cost of endoscope procurement and maintenance, coupled with the susceptibility of optical components to damage during high-temperature sterilization (potentially affecting image accuracy), necessitates regular technical maintenance. In this study, periodontal endoscopic findings were validated through extraction, flap surgery, RCT, and follow-up, thus confirming the value of this approach for diagnosing S-VRFs. Periodontal endoscopy reduced the need for flap surgery by 64.28% and reduced the need for RCT by 42.22%. Combining CBCT with periodontal endoscopy improved the diagnostic accuracy for S-VRFs by 49.1%. However, this study has several limitations. ① Patients were recruited via convenience sampling; therefore, only patients with suspected VRF who presented at the Hospital of Stomatology, Xi’an Jiaotong University, were included, thus resulting in a higher prevalence of VRFs (62.5%, 70/112) that may not fully represent the broader S-VRF population. ② The non-endodontically treated tooth (NETT) rate in this study (9.8%) was substantially lower than the 40% rate reported elsewhere [30]. This discrepancy may be due to the exclusion of NETT cases, as CBCT can visualize clear fracture lines unhindered by root fillings. ③ Most S-VRF cases in this cohort had coronal restorations, suggesting a potential association with increased VRF risk; however, the small sample size precluded confirmation of this hypothesis. ④ the follow-up duration was limited to 1 year. Ongoing sample collection and planned 1-, 3-, and 5-year follow-up assessments will be used to monitor tooth status, gather additional patient information (e.g., occlusal contact, T-SCAN data, lifestyle habits, and dietary factors) and further analyse similarities and differences between confirmed and suspected VRF cases. Additionally, future research will aim to explore the aetiology of VRF and determine whether preventive modifications to diets or occlusion could reduce the incidence of VRF. The accuracy of periodontal endoscopy for detecting VRF is influenced by the morphology of the alveolar bone defect, the location of the fracture, and the depth of the fracture. Herein, the VRFs that were missed by periodontal endoscopy but detected during flap surgery all occurred in premolars and exhibited distinct buccal bone loss on radiographic examination. The specific limitations in these cases were as follows. Case 1: The buccal alveolar bone defect presented with a narrow coronal and wide apical configuration (“top-narrow, bottom-wide”). The minimal bone loss near the cementoenamel junction hindered the insertion of the endoscope probe into the periodontal pocket. Case 1: The buccal alveolar bone defect presented with a narrow coronal and wide apical configuration (“top-narrow, bottom-wide”). The minimal bone loss near the cementoenamel junction hindered the insertion of the endoscope probe into the periodontal pocket. Case 2: The fracture originated from the buccal root axis angle and extended to the distal surface. The presence of adjacent teeth prevented adequate access for the endoscope probe to the mesial and distal aspects of the root. Case 2: The fracture originated from the buccal root axis angle and extended to the distal surface. The presence of adjacent teeth prevented adequate access for the endoscope probe to the mesial and distal aspects of the root. Case 3: The fracture line was shallow. Although repeated methylene blue staining and subsequent high-magnification microscopy revealed a faint blue-stained crack, the resolution of the endoscope probe was insufficient for clear visualization of this early-stage fracture. Case 3: The fracture line was shallow. Although repeated methylene blue staining and subsequent high-magnification microscopy revealed a faint blue-stained crack, the resolution of the endoscope probe was insufficient for clear visualization of this early-stage fracture. In this study, VRFs were more prevalent in females than in males, consistent with the findings of Pan Xiao [31]. This disparity may be accounted for by sex differences in dietary preferences, masticatory frequency, and oral parafunctional habits [32]. The mean age of our patients with VRF who had undergone RCT and experienced VRF was 46.5 years, whereas the mean age of patients without RCT was 48.56 years. This finding suggests that RCT accelerates the occurrence of VRFs, consistent with the findings of Pan et al. [31]. It is possible that the cutting of the root canal wall during RCT, the pressure during filling, and the swelling properties of the filling material act together to promote root fracture [33, 34]. In this study, periodontal endoscopy indicated high sensitivity for VRF detection, and thus was associated with a reducde need for exploratory flap surgery and RCT. Based on our findings, periodontal endoscopy could be considered an adjunct to CBCT for screening VRFs. For teeth that are diagnosed with S-VRFs via CBCT and exhibit isolated periodontal pockets, periodontal endoscopy may be performed first. Subsequent treatment (extraction, flap surgery, or RCT) could be determined on the basis of the presence or absence of an endoscopically visualized fracture. In the future, the design of periodontal endoscope probes can be optimized. Sheaths of varying lengths, widths, and angulations could be developed to adapt to diverse root morphologies. Furthermore, enhancing probe resolution is crucial for improving the detection of subtle, early-stage cracks, enabling prompt diagnosis and intervention. Integration with Artificial Intelligence (AI) and Augmented Reality (AR) technologies offers promising avenues: AR-guided navigation could provide intuitive root surface visualization, whereas AI algorithms could prioritize the examination of high-susceptibility areas, increase efficiency, shorten the procedure time, and improve the patient experience. AI-assisted diagnosis, trained on extensive libraries of periodontal endoscopic images of VRFs, warrants investigation to aid clinicians in identifying challenging fractures (e.g., apical, early-stage fractures). But their clinical efficacy requires validation in larger, multicenter studies. Conclusion Periodontal endoscopy, as an auxiliary examination method for VRF, was associated with changes in optimization of the treatment pathway for S-VRFs, prevention of overtreatment, and promotion of precision dentistry development. In the future, it is necessary to address technology-related blind spots and establish a standardized training system. Acknowledgements The experimental procedures were performed at the Key Laboratory of Shaanxi Province for Craniofacial Precision Medicine Research, Clinical Research Center of Shaanxi Province for Dental and Maxillofacial Diseases, and Department of Cariology & Endodontics, College of Stomatology, Xi’an Jiaotong University, Xi’an, China. Abbreviations Vertical root fracture Suspected vertical root fracture Cone beam computed tomography Root canal treatment Root canal retreatment Probing depth Clinical attachment loss Non-endodontic treatment Endodontically treated teeth Receiver operating characteristic Area under the curve Positive Likelihood Ratio Negative Likelihood Ratio Positive Predictive Value Negative Predictive Value Artificial Intelligence Augmented Reality Authors’ contributions All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by M Yang, YP Xue. The first draft of the manuscript was written by M Yang. QQ Dong supervised the project. All authors commented on previous versions of the manuscript. All authors read and approved the final manuscript. Funding Not applicable. Data availability No datasets were generated or analysed during the current study. Declarations Ethics approval and consent to participate This study was approved by the Institutional Review Board of Xi’an Jiaotong University Stomatological Hospital (No. KY-QX-20240004). Written informed consent was obtained from all participants, and the study was conducted in compliance with the Declaration of Helsinki (https://www.wma.net/policies-post/wma-declaration-of-helsinki/). Consent for publication All participants provided written informed consent for the publication of their personal clinical details and any identifying images included in this study. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Yunpeng Xue, Email: 41713678@qq.com. Qianqian Dong, Email: qiamqian@163.com. References Associated Data Data Availability Statement No datasets were generated or analysed during the current study.