Abstract
Objective
To assess the diagnostic performance of photon-counting detector computed tomography (PCD-CT) and cone-beam computed tomography (CBCT) at dose-matched radiation levels (high, standard, and low) for detecting and evaluating simulated endodontic conditions, treatments, and associated complications.
Methods
Sixteen extracted third molars with eight endodontic tasks were imaged using PCD-CT and CBCT. Qualitative (image quality, artifact susceptibility, diagnostic interpretability) and quantitative (endodontic working length) parameters were assessed by two observers using a five-point Likert scale. Descriptive statistics and weighted kappa (κ) were used for data analysis.
Results
High- and standard-dose PCD-CT demonstrated superior image quality and anatomical visualization compared to CBCT (median 5, IQR 5–5; κ = 1.0; all p < 0.001). Low-dose PCD-CT remained diagnostically robust, outperforming CBCT, except in root canal visualization, where both performed similarly. Diagnostic accuracy of pathologies and complications was slightly higher with PCD-CT (80–88%) than with CBCT (75–88%). Endodontic working length measurements were consistently accurate across all protocols, with near-perfect inter-observer agreement (κ = 0.84–0.86, all p < 0.001).
Conclusions
PCD-CT demonstrated superior diagnostic performance over CBCT across multiple endodontic tasks, particularly at high and standard doses. Even at low doses, PCD-CT maintained robust accuracy and image quality, outperforming dose-matched CBCT in most parameters. Endodontic working length assessment was equally reliable across both modalities. Overall, PCD-CT offers diagnostic advantages over CBCT, particularly in challenging cases involving complex anatomy or high-density materials. Its effective performance at lower radiation levels emphasizes its clinical potential and supports broader implementation in dentomaxillofacial diagnostics.
Keywords: (MeSH): endodontics; Cone-Beam computed tomography; Tomography, X-Ray computed; Radiation dosage; Diagnostic imaging
Background
Dental endodontic treatment is often required in cases of pulpal inflammation resulting from persistent microbes in cariogenic lesions, failed restorative procedures, or macro- or microtrauma [1, 2]. The treatment is essential for preventing tooth loss and involves several clinically challenging steps, including root canal preparation, chemo-mechanical debridement, and obturation [3]. Each step demands precise radiographic assessment, particularly for evaluating the preoperative periapical status and apical obturation length, to ensure long-term treatment success [4]. Additionally, complications such as root perforations, fractured instruments, or incomplete obturation may arise during treatment, compromising tooth survival. Accurate radiological visualization of these lesions is crucial for diagnosis and effective clinical management.
Conventional periapical radiographic images project three-dimensional anatomy onto a two-dimensional plane, visualizing the mesiodistal aspect of the tooth and periradicular bone. However, radiographic images are prone to diagnostic inaccuracies due to the superimposition of overlying anatomy and the density of cortical bone, particularly when evaluating the bucco-lingual axis [5, 6]. To overcome these limitations, three-dimensional (3D) imaging techniques such as cone-beam computed tomography (CBCT) have been implemented in endodontic workflows [7]. CBCT is widely established in endodontics for assessing root canal morphology, periapical pathology, anatomical variations, and treatment-related complications [8–10]. Studies have demonstrated that the diagnostic accuracy of CBCT is significantly higher than that of periapical radiographs, especially in cases involving complex endodontic conditions [7, 11]. However, CBCT is limited by its low contrast resolution and artifacts induced by high-density endodontic materials, such as gutta-percha, sealers, and metal posts [12]. Recently introduced photon-counting detector (PCD-) CT has the potential to overcome these limitations by providing a higher signal-to-noise ratio (SNR), enhanced contrast resolution, and improved radiation dose efficiency, while achieving a spatial resolution comparable to that of CBCT [13–15]. Recent findings highlight the superior performance of PCD-CT imaging in dental diagnostic applications, particularly for the visualization of dental structures [16, 17], surrounding hard and soft tissues [18], and related pathologies [14, 19–21].
Despite the increasing clinical use of PCD-CT in dentomaxillofacial radiology [22, 23], data on its application in endodontics are scarce. While previous studies have explored submillimeter visualization of dental structures, caries detection, and assessment of surrounding tissues, its role in evaluating complex endodontic conditions, treatment outcomes, and associated complications remains largely unexamined, with minimal in vivo data on diagnostic accuracy, radiation dose, or cost-effectiveness [14, 24–27]. Most current research still focuses on conventional CBCT or micro-CT, leaving a clear gap regarding the potential advantages of PCD-CT for endodontic tasks.
Therefore, this ex vivo study aimed to assess the diagnostic performance of PCD-CT and CBCT at dose-matched radiation levels (high, standard, and low) for detecting and evaluating simulated endodontic conditions, treatments, and associated complications.
The null hypothesis of this study is that there is no difference in diagnostic performance between PCD-CT and CBCT across various endodontic tasks at matched radiation doses.
Methods
Study design and ethics
This study was designed as an exploratory feasibility assessment; therefore, no formal power calculation was performed. The sample size was estimated based on the literature on established ex vivo endodontic imaging studies. Sixteen extracted human mandibular third molars were obtained from the Department of Cranio-Maxillofacial and Oral Surgery. Each specimen was checked to confirm the absence of external resorption or post-extraction damage. These teeth were then used to simulate endodontic conditions, treatments, and complications relevant to peri-endodontic treatment.
Specimen preparation
The simulated endodontic diagnostic conditions were created by an experienced dentist with ten years of clinical experience using standardized techniques and instruments. All simulations were validated through magnified visual inspection and individual periapical radiographs of each tooth to confirm accurate morphology, size, and clinical relevance.
The preparations included eight diagnostic tasks: Teeth 1 and 2 were left unaltered, serving as the control group. Teeth 3 and 4 had carious lesions. Teeth 5 and 6 had instrumented root canals that were obturated with Cavit (3 M Espe, St. Paul, USA). Teeth 7 and 8 had their root canals filled with gutta-percha (FKG Dentaire, La Chaux-de-Fonds, Switzerland). Teeth 9 and 10 simulated a fractured Hedstrom file (FKG Dentaire, La Chaux-de-Fonds, Switzerland) located in the apical third of the root. Tooth 11 exhibited a root perforation in the distal root without subsequent endodontic treatment, while tooth 12 had a similar perforation but underwent complete endodontic treatment. Tooth 13 was prepared with a cast post placed during endodontic preparation prior to obturation or cementation. This setup was intended to simulate the presence of a high-density post material during endodontic procedures, whereas tooth 14 had a cast post placed with a root perforation and received endodontic treatment. Titanium posts were used to simulate clinical scenarios involving high-density restorative materials, as they are commonly used in clinical practice. Finally, teeth 15 and 16 were prepared to simulate external root resorption. The root canal treatment was performed as follows: After access cavity preparation, patency was confirmed using a size 10 K-file (Dentsply Sirona Endodontic, Bensheim, Germany), and the working length was determined to be 1 mm short of the apical foramen. Thereafter, the root canals were enlarged using the ProTaper Profile Orifice Shapers System (Dentsply Sirona, Bensheim, Germany). The root canals were irrigated using 5 ml of 2.5% sodium hypochlorite and thereafter filled using the lateral condensation technique with gutta-percha and AH Plus resin sealer (Dentsply, Bensheim, Germany). The titanium post (Titanium (Mooser) Cendres + Métaux SA, Biel-Bienne, Switzerland) was inserted according to the manufacturer’s instructions. The simulated external root resorption was created using a 1.1 mm round diamond bur (FG 4200, Intensive, Montagnola, Switzerland).
Two silicone models were designed using Exocad DentalCAD software (Exocad GmbH, Darmstadt, Germany) and manufactured using a 3D printing machine D4K Pro Dental (Envisiontec, Dearborn, USA) and the 3D resin Keyprint KeyModel Ultra (Keyprint, Gibbstown, USA). The 3D-printed models embedded the teeth with simulated conditions and complications to replicate in vivo imaging conditions.
The dental models were then subjected to a series of imaging procedures, including CBCT and PCD-CT scans, which were performed and supervised on the same day by trained investigators (A.A.H., V.M., T.T.D.) from the Institute of Diagnostic and Interventional Radiology and the Department of Cranio-Maxillofacial and Oral Surgery.
According to the guidelines of the Local Ethics Committee (Zurich, Switzerland), no ethical approval was required. All experiments were carried out in accordance with the Declaration of Helsinki and its subsequent revisions regarding medical research.
Image acquisition
A field of view (FOV) of 13 × 12 cm was selected to ensure that all sixteen teeth were captured in a single scan, allowing consistent alignment and comparative analysis across all dose levels. Although smaller FOVs are available, reducing the FOV would have required multiple scans, increasing variability and complicating standardized comparisons.
CBCT data acquisition
The teeth were scanned using a CBCT system (NewTom 7G, QR Systems, Verona, Italy) according to the manufacturer’s recommended high, standard, and low-dose protocols. The teeth model was centrally positioned and aligned on a platform, ensuring proper orientation with the aid of the scanner’s positioning lights. The high-dose protocol employed settings of 120 kV, 11 mA, and an exposure time of 3.8 s, with a pixel size of 154 μm and a FOV of 13 × 12 cm, resulting in an effective dose of 108 µSv. The standard-dose protocol utilized 120 kV, 7 mA, and the same exposure, pixel size and FOV, yielding an effective dose of 78.4 µSv. The low-dose protocol was performed at 100 kV and 8 mA with a shortened exposure time of 1.4 s, while maintaining the same pixel size and FOV, resulting in an effective dose of 14.7 µSv.
PCD-CT data acquisition
The imaging was performed using a first-generation dual-source PCD-CT system (NAEOTOM Alpha; Siemens Healthineers AG, Forchheim, Germany, version VB20) equipped with two cadmium telluride detectors operating in the ultra-high resolution mode. Detector collimation was 120 × 0.2 mm. Scanning was conducted at a tube voltage of 140 kV with tin pre-filtration, and a pitch factor of 0.85 was applied. The radiation doses for the PCD-CT scans were adjusted to align with CBCT protocols: for the high-dose protocol, the volume CT dose index (CTDIvol) was set at 3.6 mGy, yielding a dose-length product (DLP) of 51 mGy ∙ cm and an effective dose of 102 µSv (conversion factor: 0.002 mSv∙mGy⁻¹∙cm⁻¹ [28]), for the standard-dose protocol the CTDIvol was set at 2.59 mGy, yielding a DLP of 37.5 mGy ∙ cm and an effective dose of 75 µSv; and for the low-dose protocol the CTDIvol was reduced to 0.46 mGy, resulting in a DLP of 6.57 mGy ∙ cm and an effective dose of 13.1 µSv. Images were reconstructed as polychromatic images (T3D) with a slice thickness and increment of 0.2 mm, a FOV of 130 × 130 mm2, and a matrix size of 1024 × 1024 pixels. The sharp reconstruction kernel Hr76 was utilized, combined with Quantum Iterative Reconstruction (QIR) at level 3.
Image analysis
A total of 16 teeth were evaluated, involving eight different diagnostic tasks. All image evaluations were conducted using the institutional Picture Archiving and Communication System (PACS) (Deep Unity Diagnost, v.1.1.1.2, Dedalus HealthCare, Bonn, Germany) on 2-megapixel high-resolution LCD monitors under consistent ambient lighting conditions. Observers were allowed to modify windowing and zoom settings according to their individual preferences. The assessments were performed by two observers with varying levels of expertise. Observer A (A.A.H.) is a resident in oral and maxillofacial surgery with four years of professional experience, while observer B (N.A.H.) is an attending physician with ten years of expertise, holding board certification in reconstructive dentistry as well as a Master of Advanced Studies in reconstructive and implant dentistry. Before image evaluation, both observers participated in a structured calibration session to ensure consistent application of the assessment criteria. Under the guidance of one of the main investigators, they jointly reviewed three representative training datasets that included examples of randomly selected diagnostic tasks, discussed decision rules, and resolved any uncertainties to standardize interpretation and scoring procedures. To ensure unbiased assessment, all imaging datasets were anonymized by an independent researcher and presented in randomized order. The observers were blinded to each other’s ratings as well as to the imaging modality and dose protocol (high-, standard-, and low-dose).
Qualitative measurements
Image quality was evaluated using a modified 5-point Likert scale, adapted from established literature [29], focusing on three key parameters: anatomical coverage, density and contrast, and artifact presence. Anatomical coverage, primarily determined by scan range and positioning, can also be influenced by device-specific factors such as FOV and resolution limitations, and was scored as follows: 5, optimal coverage fully meeting clinical requirements; 4, adequate visibility of clinically relevant areas; 3, potentially relevant coverage requiring further inspection; 2, suboptimal coverage and irrelevant to clinical needs; 1, non-diagnostic coverage. Anatomical coverage was assessed to account not only for the nominal scan range, but also for potential device-specific limitations such as edge distortions, reconstruction artifacts, or decreased peripheral visibility, which may be more pronounced in low-dose protocols despite the FOV remaining constant. Density and contrast were evaluated as follows: 5, excellent differentiation between enamel and dentin; 4, satisfactory differentiation between enamel and dentin; 3, unsatisfactory density but adequate contrast between enamel and dentin; 2, poor density and inadequate contrast between enamel and dentin; 1, for non-diagnostic quality. Artifact presence was scored as follows: 5, no artifacts; 4, minimal streak artifacts; 3, moderate streak artifacts; 2, significant artifact interference; 1, non-diagnostic.
The visualization of key endodontic anatomical structures, including enamel, dentine, pulp chamber, root canals, and the apical foramen, was graded as follows: 5, fine details with full diagnostic interpretability; 4, small details with good diagnostic interpretability; 3, only broad details visible, affecting diagnostic interpretability; 2, significant structures are not visible, offering no diagnostic interpretability; 1, no structures visible, with no diagnostic interpretability.
Additionally, the simulated endodontic conditions, treatments, and complications were classified as follows: 0, unidentified pathology/complication; 1, pathology/complication identified with inaccurate diagnosis; and 2, pathology/complication identified with accurate diagnosis.
Quantitative measurements
Quantitative analysis for endodontic treatment planning was conducted as follows: The endodontic working length of each tooth’s root canal was measured, defined as the distance from a specified coronal reference point to the apical terminus of canal preparation and obturation [30]. Measurements were initially performed in the sagittal plane on standard-dose CBCT images and subsequently repeated on high- and low-dose CBCT, as well as on corresponding dose-equivalent PCD-CT images. To ensure consistency, investigators were instructed to take measurements at a predefined coronal reference point, specifically the most distal and highest anatomical point within the tooth structure of the specified object. The software automatically adjusted the window width and level for the standard-dose CBCT protocol, standardizing the viewing conditions across all imaging protocols and modalities. The measurements, conducted by the same two observers, were recorded in millimeters. If a clear visualization of the tooth was impossible due to artifacts or other limitations, cases could be classified as unsuitable.
Statistical analysis
Qualitative data, including parameters such as image quality, the presence of artifacts, and diagnostic interpretability related to endodontic treatment, were analyzed using descriptive statistics. This analysis encompassed the calculation of means, standard deviations (SD), medians, minimum and maximum values, and interquartile ranges (IQR). The accuracy of the diagnostic task classification was evaluated as a percentage of correct classifications. Additionally, inter-observer agreements were analyzed using weighted kappa (κ), with an interpretation as follows: poor, < 0; slight, 0-0.2; fair, 0.21–0.4; moderate, 0.41–0.6; substantial, 0.61–0.8; almost perfect 0.81-1 [31].
For the quantitative assessment of the endodontic working length, the mean differences between observers were analyzed, and the inter-observer reliability of absolute length measurements was determined and reported as κ values.
Statistical analyses were conducted using IBM SPSS Statistics software (version 29.0.2.0, IBM, Chicago, IL, USA), with the significance threshold set at α = 0.05.
Results
A total of 16 human teeth were assessed across eight distinct diagnostic tasks and complications. These evaluations were performed at equivalent radiation doses, including high-, standard-, and low-dose CBCT and PCD-CT scans, yielding 96 evaluations per observer.
Qualitative results
Anatomy coverage scores were consistently highest for high- and standard-dose PCD-CT (median 5, IQR 5–5). Standard-dose CBCT showed lower scores (median 4, IQR 3–5 and 3.25–4.75). Among low-dose protocols, PCD-CT outperformed CBCT (median 5, IQR 4–5 vs. median 3, IQR 3–4).
Density and contrast followed a similar pattern: high- and standard-dose PCD-CT achieved median 5, low-dose PCD-CT slightly lower (median 4, IQR 4–5), while standard- and low-dose CBCT showed lower median scores (3–4) with wider IQRs (3–4 to 4–4.75).
Artifact analysis further underscored the superiority of PCD-CT, which demonstrated excellent performance and consistently outperformed dose-matched CBCT protocols across all radiation levels. Notably, although low-dose PCD-CT was slightly less effective than its higher-dose counterparts, it still demonstrated robust performance, with a median score of 4 (IQR: 4–4) (Table 1).
Table 1.
Qualitative evaluation of image quality, including anatomy coverage, density and contrast, and the presence of artifacts, was conducted using both CBCT and PCD-CT at dose-matched radiation exposure levels by two independent observers (Observer A; resident, observer B: attending): A five-point analog visual scale was used (5 = most favorable, 1 = least favorable). Values are reported as medians with interquartile ranges in parentheses. Additionally, inter-observer agreement is expressed as weighted kappa with the corresponding 95% confidence interval
| Imaging Protocol | Observer A | Observer B | Inter-observer Agreement | |
|---|---|---|---|---|
| Anatomy Coverage |
HD-CBCT SD-CBCT LD-CBCT |
4 (3.25-5) 4 (3–5) 3 (3–4) |
4 (4–5) 4 (3.25–4.75) 3 (3–4) |
0.7 (0.42–0.97); p < 0.001 0.72 (0.5–0.95); p < 0.001 0.73 (0.45–1.02); p < 0.001 |
|
HD-PCD-CT SD-PCD-CT LD-PCD-CT |
5 (5–5) 5 (5–5) 5 (4–5) |
5 (5–5) 5 (5–5) 5 (4–5) |
1 (1–1); p < 0.001 1 (1–1); p < 0.001 0.862 (0.6–1.12); p < 0.001 |
|
| Density and Contrast |
HD-CBCT SD-CBCT LD-CBCT |
5 (4–5) 4 (4–5) 4 (3–4) |
5 (4–5) 4 (4-4.75) 3 (3–4) |
0.73 (0.43–1.04); p < 0.001 0.81 (0.56–1.06); p < 0.001 0.65 (0.38–0.93); p < 0.001 |
|
HD-PCD-CT SD-PCD-CT LD-PCD-CT |
5 (5–5) 5 (5–5) 4 (4–5) |
5 (5–5) 5 (5–5) 4 (4–5) |
1 (1–1); p < 0.001 1 (1–1); p < 0.001 0.71 (0.42-1); p < 0.001 |
|
| Artifacts |
HD-CBCT SD-CBCT LD-CBCT |
4.5 (3–5) 4 (3.25-5) 3 (3–4) |
4 (3.25-5) 4 (4–5) 3 (3–3) |
0.78 (0.56–0.99); p < 0.001 0.67 (0.38–0.96); p < 0.001 0.72 (0.43-1); p < 0.001 |
|
HD-PCD-CT SD-PCD-CT LD-PCD-CT |
5 (5–5) 5 (5–5) 4 (4–4) |
5 (5–5) 5 (5–5) 4 (4–4) |
1 (1–1); p < 0.001 1 (1–1); p < 0.001 0.65 (0.38–0.93); p < 0.001 |
CBCT Cone-Beam Computed Tomography,PCD-CT Photon-Counting Detector Computed Tomography, HD High-Dose, SD Standard-Dose, LD Low-Dose
Assessment of key endodontic anatomical structures showed highest visualization scores for high- and standard-dose PCD-CT. Low-dose PCD-CT generally outperformed dose-matched CBCT, except for root canal visualization, where scores were similar.
For further details on the evaluation of key endodontic anatomical structures, refer to Table 2. The corresponding distribution of ordinal visual grading scores is illustrated in Fig. 1.
Table 2.
The radiographic evaluation of key endodontic anatomical structures, including enamel, dentin, pulp chamber, root canals, and the apical foramen, was performed independently by two observers based on a five-point analog visual scale (5 = most favorable, 1 = least favorable). Values are reported as medians with interquartile ranges in parentheses. Additionally, for each imaging modality (cone-beam computed tomography (CBCT) and photon-counting detector computed tomography (PCD-CT)) and their respective imaging protocols (high-, standard-, and low-dose), inter-observer agreement is quantified using weighted kappa statistics with corresponding 95% confidence interval
| Imaging Protocol | Observer A | Observer B | Inter-observer Agreement | |
|---|---|---|---|---|
| Enamel |
HD-CBCT SD-CBCT LD-CBCT |
5 (4–5) 4 (4–4) 3 (3–4) |
5 (4–5) 4 (4-4.75) 3 (3–3) |
0.72 (0.4–1.04); p < 0.001 0.79 (0.54–1.06); p < 0.001 0.74 (0.47–1.01); p < 0.001 |
|
HD-PCD-CT SD-PCD-CT LD-PCD-CT |
5 (5–5) 5 (5–5) 4 (4–4) |
5 (5–5) 5 (5–5) 4 (4–4) |
1 (1–1); p < 0.001 1 (1–1); p < 0.001 0.65 (0.38–0.93); p < 0.001 |
|
| Dentine |
HD-CBCT SD-CBCT LD-CBCT |
4 (4–5) 4 (4–4) 3 (3–3) |
4 (4–5) 4 (4-4.75) 3 (3–3) |
0.72 (0.41–1.03); p < 0.001 0.6 (0.25–0.94); p < 0.001 0.71 (0.37–1.05); p < 0.001 |
|
HD-PCD-CT SD-PCD-CT LD-PCD-CT |
5 (5–5) 5 (5–5) 4 (3.25-4) |
5 (5–5) 5 (5–5) 4 (3–4) |
1 (1–1); p < 0.001 1 (1–1); p < 0.001 0.85 (0.56–1.14); p < 0.001 |
|
| Pulp Chamber |
HD-CBCT SD-CBCT LD-CBCT |
4.5 (4–5) 4 (4-4.75) 3 (3-3.75) |
5 (4–5) 4 (3.75–4.75) 3 (3–3) |
0.73 (0.44–1.03); p < 0.001 0.73 (0.44–1.03); p < 0.001 0.71 (0.4–1.02); p < 0.001 |
|
HD-PCD-CT SD-PCD-CT LD-PCD-CT |
5 (5–5) 5 (5–5) 4 (4–4) |
5 (5–5) 5 (5–5) 4 (4–4) |
1 (1–1); p < 0.001 1 (1–1); p < 0.001 0.79 (0.54–1.06); p < 0.001 |
|
| Root Canals |
HD-CBCT SD-CBCT LD-CBCT |
5 (5–5) 5 (5–5) 4 (4–4) |
5 (5–5) 5 (5–5) 4 (4–4) |
1 (1–1); p < 0.001 1 (1–1); p < 0.001 0.7 (0.4–0.98); p < 0.001 |
|
HD-PCD-CT SD-PCD-CT LD-PCD-CT |
5 (5–5) 5 (5–5) 4 (4–4) |
5 (5–5) 5 (5–5) 4 (4–4) |
1 (1–1); p < 0.001 1 (1–1); p < 0.001 1 (1–1); p < 0.001 |
|
| Apical Foramen |
HD-CBCT SD-CBCT LD-CBCT |
5 (3.25-5) 4 (3.25-5) 3 (3-3.75) |
4.5 (4–5) 4 (3-4.75) 3 (3–3) |
0.68 (0.43–0.92); p < 0.001 0.77 (0.54–1.01); p < 0.001 0.71 (0.40–1.02); p < 0.001 |
|
HD-PCD-CT SD-PCD-CT LD-PCD-CT |
5 (5–5) 5 (5–5) 4 (3–4) |
5 (5–5) 5 (5–5) 3 (3–4) |
1 (1–1); p < 0.001 1 (1–1); p < 0.001 0.66 (0.28–1.04); p < 0.001 |
CBCTCone-Beam Computed Tomography, PCD-CT Photon-Counting Detector Computed Tomography, HD High-Dose, SD Standard-Dose, LD Low-Dose
Fig. 1.
Distribution of ordinal visual grading scores from qualitative assessments of key endodontic anatomical structures, including enamel, dentin, pulp chamber, root canals, and apical foramen. Scores are presented for each imaging modality with its corresponding dose-matched protocols (scale: 5 = best visualization, 1 = poorest)
Inter-observer agreement was consistently perfect for anatomy coverage, density and contrast, artifact susceptibility, and the depiction of dental and mandibular anatomical structures, using high- and standard-dose PCD-CT protocols (κ = 1, all p < 0.001). The agreement remained substantial to almost perfect for all low-dose PCD-CT protocols (κ = 0.65-1, all p < 0.001), highlighting the diagnostic robustness of PCD-CTs even at reduced radiation doses. In contrast, CBCT demonstrated greater variability across all evaluated parameters, with κ values ranging from 0.6 to 1.00 (all p < 0.001), showing a tendency toward slightly lower inter-observer agreement overall (Tables 1 and 2).
Endodontic classification, including the accurate diagnosis of pathologies and complications, was comparable to, or slightly higher with PCD-CT, achieving an average accuracy rate of 80–88%. In contrast, CBCT protocols showed lower accuracy at low dose (75–81%) and comparable accuracy at standard and high doses (81–88%). The most frequent difficulties in accurate classification were related to root perforation, especially in PCD-CT, and the recognition of external resorption. Unidentified pathology or complications were observed at a similar rate of approximately 5% in both modalities. The inter-observer agreement across both modalities was substantial to almost perfect, with κ values ranging from 0.62 to 1, and significant p-values in both modalities (Figs. 2 and 3; Table 3).
Fig. 2.
Examples for the visualization of simulated endodontic conditions across various diagnostic tasks using photon-counting detector computed tomography (PCD-CT) and cone-beam computed tomography (CBCT) at dose-matched high, standard, and low radiation levels. A-F: Control group; G-L: Carious lesions; M-R: Instrumented root canals obturated with Cavit; S-X: Root canals filled with gutta-percha
Fig. 3.
Comparative visualization of examples for simulated endodontic conditions across distinct diagnostic tasks using dose-matched photon-counting detector computed tomography (PCD-CT) and cone-beam computed tomography (CBCT). A-F: Fractured Hedstrom file located in the apical third of the root; G-L: Root perforation in the distal root without subsequent endodontic treatment; M-R: Cast post placed with a root perforation with endodontic treatment; S-X: External cervical root resorption
Table 3.
Classification of simulated endodontic conditions and complications (N = 16). Pathologies/complications were classified as follows: 2 = correct classification (accurate diagnosis), 1 = inaccurate diagnosis, and 0 = unidentified pathology/complication. The table reports the percentage of correct classifications for the endodontic diagnostic task and the inter-observer agreement (weighted kappa) with 95% confidence intervals
| Imaging Protocol | Observer A | Observer B | Inter-observer Agreement | |
|---|---|---|---|---|
| Endodontic Classification |
HD-CBCT SD-CBCT LD-CBCT |
88% 88% 81% |
81% 81% 75% |
0.62 (0.05–1.19); p = 0.002 0.62 (0.33–1.19); p = 0.002 0.62 (0.33–1.19); p = 0.002 |
|
HD-PCD-CT SD-PCD-CT LD-PCD-CT |
88% 88% 88% |
88% 81% 81% |
1 (1–1); p < 0.001 0.62 (0.05–1.19); p = 0.002 0.62 (0.05–1.19); p = 0.002 |
CBCT Cone-Beam Computed Tomography, PCD-CT Photon-Counting Detector Computed Tomography, HD High-Dose, SD Standard-Dose, LD Low-Dose, Min. Minimum, Max. Maximum
Quantitative results
For quantitative analysis, all protocols and modalities consistently enabled accurate endodontic work length measurements. Absolute length measurements demonstrated high inter-observer reproducibility across modalities, with weighted κ values ranging from 0.84 to 0.86 for PCD-CT and 0.82 to 0.86 for CBCT (all p < 0.001), indicating almost perfect agreement. The mean differences between the two raters were minimal across all imaging protocols, irrespective of the observer’s specialty or experience (Table 4).
Table 4.
Mean inter-observer differences in endodontic working length measurements are reported in millimeters, including standard deviations (SD), minimum and maximum values. Inter-observer reliability is expressed using weighted kappa statistics with corresponding 95% confidence intervals
| Observer Pairs | Imaging Protocol | Mean (mm) |
SD | Min. (mm) |
Max. (mm) |
Inter-observer Agreement |
|---|---|---|---|---|---|---|
| Endodontic working length |
HD-CBCT SD-CBCT LD-CBCT |
0.02 − 0.01 0.002 |
0.03 0.01 0.02 |
− 0.03 − 0.04 − 0.03 |
0.08 0.003 0.05 |
0.82 (0.74–0.89); p < 0.001 0.82 (0.72–0.92); p < 0.001 0.86 (0.77–0.95); p < 0.001 |
|
HD-PCD-CT SD-PCD-CT LD-PCD-CT |
− 0.002 − 0.001 − 0.004 |
0.03 0.02 0.02 |
− 0.09 − 0.05 − 0.08 |
0.08 0.05 0.02 |
0.84 (0.73–0.95); p < 0.001 0.86 (0.77–0.95); p < 0.001 0.84 (0.70–0.98); p < 0.001 |
CBCT Cone-Beam Computed Tomography, PCD-CT Photon-Counting Detector Computed Tomography, HD High-Dose, SD Standard-Dose, LD Low-Dose, Min. Minimum, Max. Maximum
Discussion
The growing reliance on image-based planning in dental procedures supports today’s multidisciplinary approach to individualized patient care, leading to considerable advancements in prevention, rehabilitation, and decision-making across all dental specialties, particularly during endodontic treatment. A key factor in the long-term success of endodontic treatments presents in precise visualization of anatomical structures and pathological conditions throughout the treatment procedure. The findings of this study indicated that PCD-CT performance is either superior or comparable in most qualitative and quantitative parameters, compared to CBCT, particularly in complex endodontic cases where artifacts from high-density materials like gutta-percha, sealers, and metal posts are present.
The most critical factor for the success of endodontic treatments is the accurate determination of the correct working length. Regarding this quantitatively evaluated parameter, all protocols and modalities consistently enabled accurate endodontic working length measurements, suggesting that PCD-CT can be used as a viable alternative to CBCT. In terms of anatomy coverage, density and contrast, as well as artifact analysis, the performance of PCD-CT was superior compared to its dose-matched counterparts. This advantage is likely attributable to its superior contrast resolution and volume coverage speed, which results from device-specific technical parameters, and reduced metal artifacts when ultra-high resolution images are used [32].
In this study, PCD-CT outperformed CBCT visualization, which contrasts with a previous study that evaluated the feasibility of PCD-CT for endodontic diagnostic tasks, where PCD-CT performed comparably to CBCT in visualizing endodontic structures [24]. Spatial resolution is thereby a crucial factor in the evaluation of 3D imaging techniques for endodontic diagnostics, as it significantly influences the accuracy and reliability of assessments. These advantages may be further influenced by technical factors such as voxel size, and the use of QIR, which can enhance spatial resolution and reduce artifacts, thereby improving diagnostic accuracy compared to conventional CBCT [24–26]. Additionally, the selection of CBCT devices might contribute to the divergent findings compared with those of Fontenele et al. [24]. While both studies [14, 24] utilized the same PCD-CT (NAETOM Alpha) device, the prior study selected the CBCT systems 3D Accuitomo 170 (J. Morita MFG. Corp., Kyoto, Japan) and NewTom VGi evo (NewTom, QR s.r.l., Verona, Italy), whereas this study employed the NewTom 7G scanner (QR Systems, Verona, Italy). The performance of HR 3D Accuitomo might be better due to the smaller voxel size and field of view, which likely results in a higher signal-to-noise ratio [24].
The greatest challenges in accurate classification for PCD-CT were related to the detection of root perforation and the identification of external resorption. The reduced detectability of root perforations in PCD-CT may be attributed to modality-specific noise patterns and image characteristics that can obscure small, low-contrast defects, thereby complicating the visualization of subtle perforations. Existing literature reports that CBCT benefits from isotropic voxels and high spatial resolution [33, 34], and that PCD-CT excels in noise reduction and contrast resolution [27]. Nonetheless, both modalities may still exhibit limitations in delineating intricate endodontic structures [24]. Image degradation due to photon starvation may also contribute, particularly in high-density regions such as the tooth apex. Nevertheless, in our study setting, PCD-CT outperformed CBCT, especially in endodontic tasks, where artifacts from high-density materials, such as gutta-percha, sealers, and metal posts, are present. Thus, based on the obtained results, simulated endodontic conditions and associated complications can be accurately and reproducibly classified using both modalities at all radiation doses. However, while Micro-CT remains the ex vivo gold standard for ultra-high-resolution imaging of dental structures, its clinical applicability is limited by prolonged scan times and high radiation doses. In contrast, PCD-CT offers high spatial resolution comparable to CBCT, while maintaining clinically acceptable radiation exposure and rapid acquisition times, making it suitable for in vivo endodontic diagnostics.
In clinical practice, initial assessment with standard radiographs, preferably intraoral radiographs, remains necessary to identify complex endodontic cases. In such scenarios, additional CBCT imaging is often required when intraoral or panoramic radiographs are insufficient. Our findings indicate that PCD-CT reliably visualizes complex endodontic anatomy, treatment outcomes, and associated complications, highlighting its potential as a practical, patient-applicable imaging modality that provides high-quality diagnostic information while offering the advantage of reduced radiation exposure. However, the comparatively high acquisition cost and limited availability of PCD-CT may currently restrict its routine use in endodontic practice. Further analyses of cost-effectiveness, taking into account the improved benefit-to-risk ratios of this technology for patients, are therefore warranted to assess its clinical potential.
Several study limitations must be considered. First, the investigation was conducted on a limited set of sixteen teeth, which inherently constrains the extent to which the outcomes can be extrapolated. As the study was primarily designed to assess feasibility, a formal power analysis was not conducted. Although ex vivo studies have successfully simulated root cracks and fractures [35], these conditions were not included in the present study due to the technical challenges of reliably reproducing such fine defects without risking uncontrolled damage to the teeth. Further research involving larger sample sizes, a broader spectrum of pathologies, and calculated statistical power is needed to validate these results. Second, the ex vivo nature of this study does not fully replicate the in vivo conditions found in humans, such as differences in tissue composition and the potential impact of patient movement. Third, although both observers were experienced clinicians, using only two evaluators may limit the generalizability and reproducibility of the findings, underscoring the need for future studies to include a larger, more diverse cohort of observers. Fourth, only one CBCT device was evaluated, highlighting the need for further validation using scanners from various vendors in larger cohorts. To validate these results, additional research in clinical settings with a diverse patient population is necessary to assess the practical use of PCD-CT in endodontics.
Conclusion
The findings of this study demonstrate the clinical potential of PCD-CT for visualizing complex endodontic conditions, treatments, and associated complications, particularly with high- and standard-dose protocols. Notably, low-dose PCD-CT also showed diagnostic value in selected cases, offering a more favorable benefit-to-risk profile by significantly reducing radiation exposure without substantially compromising diagnostic accuracy. Furthermore, endodontic working length measurements were equally accurate across all radiation dose levels for both PCD-CT and CBCT protocols.
Acknowledgements
The authors thank Debora Putignano for her support during the image acquisition. They also thank Kiren Mätzener, Michele Piccini, and Mutlu Özcan for their valuable assistance in the realization of the in vitro human tooth model holder.
Financial Interests
VM, HA, TD, EB, TF: Institutional grants from Bayer, Canon, Guerbet, and Siemens. TF is a retired employee of Siemens.
Authors’ contributions
Conceptualization, methodology, execution, or formal analysis, A.A.H., V.M., N.A.H, H.A., T.T.D., H.E., S.W., T.F., S.V., S.D., E.B., B.S.; drafting manuscript, A.A.H., writing review and editing, A.A.H., V.M., N.A.H, H.A., T.T.D., H.E., S.W., T.F., S.V., S.D., E.B., B.S. All authors have read and agreed to the final version of the manuscript. All authors agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
Funding
No funding.
Data availability
The datasets used during and/or analyzed during the current study are available from the corresponding author on request.
Declarations
Ethics approval and consent to participate
This study was not submitted to or reviewed by an ethics committee, as ethical approval was not required according to national regulations and the guidelines of the Cantonal Ethics Committee Zurich (Kantonale Ethikkommission Zürich, Switzerland). All experiments were conducted in accordance with the Declaration of Helsinki and its subsequent revisions.
Consent for publication
not applicable.
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.
References
- 1.Farges JC, Alliot-Licht B, Renard E, Ducret M, Gaudin A, Smith AJ, et al. Dental pulp defence and repair mechanisms in dental caries. Mediators Inflamm. 2015;2015:230251. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Marchionatti AME, Wandscher VF, Rippe MP, Kaizer OB, Valandro LF. Clinical performance and failure modes of pulpless teeth restored with posts: a systematic review. Braz Oral Res. 2017;31:e64. [DOI] [PubMed] [Google Scholar]
- 3.Jakovljevic A, Fransson H, Bakhsh A, Jacimovic J, Lazic EK, Ivanovic KB, et al. Endodontic treatment modifies circulatory inflammatory mediator levels: A systematic review with meta-analysis. Int Endod J. 2025;58(2):171–92. [DOI] [PubMed] [Google Scholar]
- 4.Basmadjian-Charles CL, Farge P, Bourgeois DM, Lebrun T. Factors influencing the long-term results of endodontic treatment: a review of the literature. Int Dent J. 2002;52(2):81–6. [DOI] [PubMed] [Google Scholar]
- 5.Patel S, Dawood A, Mannocci F, Wilson R, Pitt Ford T. Detection of periapical bone defects in human jaws using cone beam computed tomography and intraoral radiography. Int Endod J. 2009;42(6):507–15. [DOI] [PubMed] [Google Scholar]
- 6.Tsurumachi T, Honda K. A new cone beam computerized tomography system for use in endodontic surgery. Int Endod J. 2007;40(3):224–32. [DOI] [PubMed] [Google Scholar]
- 7.Cotton TP, Geisler TM, Holden DT, Schwartz SA, Schindler WG. Endodontic applications of cone-beam volumetric tomography. J Endod. 2007;33(9):1121–32. [DOI] [PubMed] [Google Scholar]
- 8.Patel S, Brown J, Pimentel T, Kelly RD, Abella F, Durack C. Cone beam computed tomography in Endodontics - a review of the literature. Int Endod J. 2019;52(8):1138–52. [DOI] [PubMed] [Google Scholar]
- 9.Magat G, Hatipoğlu Ö, Köse T, Hatipoğlu FP. Root Canal morphology of mandibular anterior permanent teeth in Turkish sub-population using two classification systems: a cone-beam computed tomography study. Odontology. 2025;113(2):799–808. [DOI] [PubMed] [Google Scholar]
- 10.Al-Haddad A, Alarami N, Alshammari AF, Hatipoğlu Ö, Hatipoğlu FP, Madfa AA. The worldwide prevalence of radix molaris in mandibular second molars assessed by cone beam computed tomography: systematic review and Meta-analysis. J Endod. 2025. [DOI] [PubMed]
- 11.Estrela C, Bueno MR, Leles CR, Azevedo B, Azevedo JR. Accuracy of cone beam computed tomography and panoramic and periapical radiography for detection of apical periodontitis. J Endod. 2008;34(3):273–9. [DOI] [PubMed] [Google Scholar]
- 12.Gaêta-Araujo H, Nascimento EHL, Fontenele RC, Mancini AXM, Freitas DQ, Oliveira-Santos C. Magnitude of beam-hardening artifacts produced by gutta-percha and metal posts on cone-beam computed tomography with varying tube current. Imaging Sci Dent. 2020;50(1):1–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Rajendran K, Petersilka M, Henning A, Shanblatt E, Marsh J, Thorne J, et al. Full field-of-view, high-resolution, photon-counting detector CT: technical assessment and initial patient experience. Phys Med Biol. 2021;66(20). 10.1088/1361-6560/ac155e. [DOI] [PMC free article] [PubMed]
- 14.Ruetters M, Sen S, Gehrig H, Bruckner T, Kim TS, Lux CJ, et al. Dental imaging using an ultra-high resolution photon-counting CT system. Sci Rep. 2022;12(1):7125. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Mergen V, Sartoretti T, Baer-Beck M, Schmidt B, Petersilka M, Wildberger JE, et al. Ultra-High-Resolution coronary CT angiography with Photon-Counting detector CT: feasibility and image characterization. Invest Radiol. 2022;57(12):780–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Al-Haj Husain A, Benz F, Mergen V, Valdec S, Winklhofer S, Alkadhi H, et al. Exploring dental anatomy: an ex vivo comparative study between Photon-Counting detector CT and Cone-Beam CT. Clin Exp Dent Res. 2025;11(6):e70266. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Stadlinger B, Engel K, Al-Haj Husain A, Alkadhi H. Photon-counting detector computed tomography Meets orthopantomography: a cinematic rendering solution for superior structural visualization. Int J Oral Maxillofac Surg. 2025;6:S0901-5027(25)01400-6. [DOI] [PubMed]
- 18.Al-Haj Husain A, Mergen V, Valdec S, Winklhofer S, Alkadhi H, Essig H, et al. Cone-beam versus photon-counting detector CT: influence of dose variations on the detection of simulated mandibular osseous lesions. J Craniomaxillofac Surg. 2025;53(8):1236–45. [DOI] [PubMed] [Google Scholar]
- 19.Al-Haj Husain A, Mergen V, Valdec S, Al-Haj Husain N, Stadlinger B, Essig H, et al. Comparison of cone-beam computed tomography with photon-counting detector computed tomography for dental implant surgery. Int J Implant Dent. 2025;11(1):21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Al-Haj Husain A, Mergen V, Sandhu S, Wagner MEH, Demmert TT, Alkadhi H, et al. Postoperative assessment of fracture reduction and osteosynthesis materials using photon-counting detector CT in maxillofacial trauma - a pilot study. Oral Maxillofac Surg. 2025;29(1):171. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Sandhu S, Al-Haj Husain A, Mergen V, Demmert TT, Alkadhi H, Flohr T, et al. Comparative evaluation of photon-counting detector CT and cone-beam CT in the assessment of simulated mandibular trauma. J Stomatol Oral Maxillofac Surg. 2025;127(2):102603. [DOI] [PubMed] [Google Scholar]
- 22.Dogra S, Shekhrajka N, Moonis G. A review of applications of photon-counting computed tomography in head and neck imaging. Br J Radiol. 2025;98(1175):1802–12. [DOI] [PubMed] [Google Scholar]
- 23.Zanon C, Pepe A, Cademartiri F, Bini C, Maffei E, Quaia E et al. Potential benefits of Photon-Counting CT in dental imaging: A narrative review. J Clin Med. 2024;13(8):2436. [DOI] [PMC free article] [PubMed]
- 24.Fontenele RC, Picoli FF, Pinto JC, Coudyzer W, de Faria Vasconcelos K, Farias Gomes A, et al. Feasibility of photon-counting computed tomography as a novel imaging modality for challenging endodontic diagnostic tasks. Sci Rep. 2023;13(1):6266. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Rau S, Pichotka MP, Rau A, Reisert M, Altenburger MJ, Schmelzeisen R, et al. Photon-counting-detector CT outperforms state-of-the-art cone-beam and energy-integrated-detector CT in delineation of dental root canals. Sci Rep. 2025;15(1):2978. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Rau S, Rau A, Lang F, Fink A, Weiß J, Kellner E, et al. Photon-counting CT outperforms dental CBCT in detecting small accessory canals in root-filled teeth in a Phantom study. Sci Rep. 2025;15(1):37352. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Vanden Broeke L, Grillon M, Yeung AWK, Wu W, Tanaka R, Vardhanabhuti V. Feasibility of photon-counting spectral CT in dental applications-a comparative qualitative analysis. BDJ Open. 2021;7(1):4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Romanyukha A, Folio L, Lamart S, Simon SL, Lee C, BODY, SIZE-SPECIFIC EFFECTIVE, DOSE CONVERSION COEFFICIENTS FOR CT SCANS. Radiat Prot Dosimetry. 2016;172(4):428–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Sabarudin A, Tiau YJ. Image quality assessment in panoramic dental radiography: a comparative study between conventional and digital systems. Quant Imaging Med Surg. 2013;3(1):43–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Ricucci D. Apical limit of root Canal instrumentation and obturation, part 1. Literature review. Int Endod J. 1998;31(6):384–93. [DOI] [PubMed] [Google Scholar]
- 31.Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1977;33(1):159–74. [PubMed] [Google Scholar]
- 32.Sawall S, Maier J, Sen S, Gehrig H, Kim TS, Schlemmer HP, et al. Dental imaging in clinical photon-counting CT at a quarter of DVT dose. J Dent. 2024;142:104859. [DOI] [PubMed] [Google Scholar]
- 33.Brüllmann D, Schulze RK. Spatial resolution in CBCT machines for dental/maxillofacial applications-what do we know today? Dentomaxillofac Radiol. 2015;44(1):20140204. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Michetti J, Maret D, Mallet JP, Diemer F. Validation of cone beam computed tomography as a tool to explore root Canal anatomy. J Endod. 2010;36(7):1187–90. [DOI] [PubMed] [Google Scholar]
- 35.Neves FS, Freitas DQ, Campos PS, Ekestubbe A, Lofthag-Hansen S. Evaluation of cone-beam computed tomography in the diagnosis of vertical root fractures: the influence of imaging modes and root Canal materials. J Endod. 2014;40(10):1530–6. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets used during and/or analyzed during the current study are available from the corresponding author on request.