上颌乳磨牙根管和根管形态的微 CT 分析
Micro-computed tomography analysis of root and canal morphology in maxillary primary molars
一项体外研究利用微 CT 技术分析了 60 颗上颌乳磨牙的三维根管解剖结构,重点评估 MB2 通道的发生率、位置及形态。该研究纳入了 41 颗第一乳磨牙和 19 颗第二乳磨牙,旨在通过高分辨率非破坏性成像,详细表征乳牙复杂的内部解剖特征。研究结果强调,准确识别 MB2 等变异结构对于提高乳牙根管治疗的成功率至关重要。
Introduction
Effective management of pulpal disease in primary teeth requires the paediatric dentist to understand both pulp biology and root canal anatomy1. In the primary dentition, the root canals often show multiple branchings, lateral canals, and apical ramifications that are difficult to detect on conventional radiographs. A three-dimensional understanding of this anatomy is therefore needed to interpret the internal structure correctly and to improve treatment outcomes in primary teeth2.
Maxillary primary molars have particularly variable and complex internal anatomy3,4. They usually present with two to four roots; three roots are the most common configuration, whereas the two-rooted variant typically results from fusion of the distobuccal and palatal roots. The mesiobuccal (MB) root is especially challenging because it frequently contains a second canal (MB2) alongside the main mesiobuccal canal (MB1). MB2 is often narrower and more tortuous than MB1, which makes it harder to locate and instrument. Reported MB2 prevalence in primary maxillary molars ranges from about 20% to 60%, and is higher in second than in first primary molars1,3. Failure to find and treat MB2 has been linked to root canal treatment failure and persistent periapical pathology5.
Because of the considerable variability in root and canal morphology, several authors have proposed systematic anatomical descriptions and classifications6,7. A few reports have documented a third mesiobuccal canal or other atypical configurations in the distobuccal and palatal roots, which reinforces the need for careful clinical and radiographic assessment8. Other recognised variations include accessory canals, particularly in the furcation area9,10, and isthmuses, the narrow pulp-filled connections between adjacent canals such as MB1 and MB211. These features hinder cleaning, shaping, and obturation, and may harbour persistent microbial biofilms that contribute to chronic apical periodontitis12.
Root canal morphology in primary teeth has been studied in vitro by several methods, including digital radiography, sectioning, the staining-and-clearing technique, scanning electron microscopy, cone-beam computed tomography (CBCT), and micro-computed tomography (micro-CT)1,2. With voxel sizes typically between 5 and 50 μm, micro-CT provides a volumetric resolution one to several orders of magnitude higher than conventional CT1. Because it is non-destructive, high-resolution, and reproducible, micro-CT allows accurate three-dimensional evaluation of canal anatomy without altering the specimen, which makes it well suited for laboratory studies .
Nevertheless, high-resolution micro-CT data on the precise location and morphometry of the MB2 canal in maxillary primary molars remain scarce. The present study therefore aimed to characterise the three-dimensional internal root canal anatomy of maxillary primary molars in detail, with particular emphasis on the prevalence, location, and morphometry of the MB2 canal, together with inter-orifice distances, canal diameters, and dentin wall thickness.
Methods
Ethics approval
The study protocol was approved by the Research Ethics Committee, Faculty of Dental Medicine for Girls, Al-Azhar University (REC-PD-26-06), and the study was conducted in accordance with the Declaration of Helsinki. Because the teeth had been extracted as part of standard clinical care for reasons unrelated to this study and were fully anonymised before transfer to the research team, the ethics committee waived the requirement for individual informed consent. No identifiable personal data are included in this manuscript; therefore, consent for publication was not applicable.
Sample size calculation
The sample size was calculated in Epi Info version 7.2.5.0 (CDC, Atlanta, USA), based on the 6.25% prevalence of the four-canal configuration (three roots with MB1, MB2, DB, and palatal canals) in primary maxillary first molars reported by Datta et al.3. With a 90% confidence interval, a 5% margin of error, and a finite population correction, the minimum required sample was 60 teeth. All 60 collected teeth met the inclusion criteria after micro-CT review and were retained for analysis. A 90% confidence level was adopted a priori because the primary aim was descriptive characterisation of anatomical prevalence rather than formal hypothesis testing, balancing estimate precision against the limited availability of intact, non-pathological extracted primary molars.
Selection criteria
Sixty extracted primary maxillary molars were collected after routine extractions performed for reasons unrelated to this study. Teeth were included if they were primary maxillary first or second molars with no more than minimal physiological root resorption (≤ 2 mm). Teeth were excluded for root fractures, internal or external pathological resorption, canal obliteration, extensive coronal or proximal restorations, previous endodontic treatment, or any evident developmental anomaly. Physiological root resorption was assessed by direct visual (naked-eye) examination of each extracted tooth during selection, and teeth showing more than 2 mm of resorption were excluded. The extent of resorption was estimated on each tooth by comparing the residual root length with the expected intact root length for that tooth type. The sample comprised 41 primary maxillary first molars (tooth type D) and 19 s molars (tooth type E). The patients were aged approximately 4–7 years, corresponding to the interval between completion of root formation and the onset of normal physiological root resorption, and the teeth had been extracted because of extensive caries with a clinical recommendation for extraction.
Sample preparation
Each specimen received a sequential alphanumeric code generated from a computer-based random number list (www.random.org) so that the examiner remained blinded to specimen identity during analysis. The teeth were washed under running water, and any remaining soft tissue was gently removed from the root surface. The specimens were stored in sterile saline at room temperature until scanning.
Micro-computed tomography acquisition and reconstruction
The teeth were scanned in a desktop SkyScan 1172 micro-CT system (Bruker microCT, Kontich, Belgium). Each specimen was mounted in a custom holder to ensure reproducible alignment. The microfocus X-ray tube operated at 85 kV and 118 µA, with a 13.61 μm isotropic reconstructed voxel size (9 μm camera pixel with 2 × 2 binning), a rotation step of 0.60°, and an exposure time of 2600 ms. Raw projections were reconstructed in NRecon (v1.6.10.6; Bruker microCT) with 49% beam-hardening correction and a ring-artefact reduction setting of 7 (software scale 0–20). Quantitative analysis was performed in CTAn (v1.15.4) and visualisation in CTVol (v2.0). The measurement protocol and the anatomical reference levels are illustrated in Figs. 1 and 2. Additional acquisition and reconstruction parameters were: an aluminium–copper (Al + Cu) filter, two-frame averaging, and random specimen movement, with flat-field, geometrical, and median-filter corrections enabled; projections were acquired over a 192.6° angular range (partial-scan mode; “use 180°+” off) in step-and-shoot mode. During NRecon reconstruction a smoothing level of 2 (asymmetrical boxcar kernel) and a Hamming reconstruction filter were applied. For qualitative visualisation and canal identification only, the reconstructed images were enhanced using a high-boost (sharpening) filter in onDemand3D software (Cybermed); all quantitative measurements were performed on the reconstructed data in CTAn. Cross-sectional measurements were made on the images in their original reconstructed orientation, and the reconstructed volumes were subsequently reoriented solely to visualise each root canal individually.
Micro-CT images illustrating the measurement protocol. (A) Coronal section showing the canal orifice level, canal length, and the three root-third levels (coronal, middle, apical); the standardised sections used for analysis were a coronal section (1 mm below the orifice), a mid-root section, and an apical section (1 mm coronal to the apical constriction). (B) Cross-section showing inter-canal distance measurements between the MB1, MB2, DB, and palatal canal centres. (C) Cross-section showing dentin thickness measurements on the buccal, lingual, mesial, and distal aspects. (D) Cross-section showing mesiodistal and buccolingual canal diameter measurements through the canal centre. Scale bar = 1 mm in all panels. All measurements were made at the calibrated isotropic voxel size of 13.61 μm using CTAn (v1.15.4).
Representative micro-CT images of a maxillary primary molar (Specimen 12) showing the imaging protocol and the anatomical reference levels. (A) Three-dimensional volume rendering, occlusal view, showing the four canal orifices on the pulp chamber floor: MB1, MB2, DB, and palatal (P). (B) Three-dimensional volume rendering, buccal view, showing the trifurcated external root morphology with the mesiobuccal (MB), distobuccal (DB), and palatal (P) roots. (C) Axial cross-section at the coronal third (1 mm below the orifice level), showing the arrangement of the canals on the pulp chamber floor. (D) Axial cross-section at the middle third, showing the distobuccal root with two independent canals (DB1, DB2), the palatal (P) canal with an oval cross-section, and the mesiobuccal (MB) root. (E) Axial cross-section at the apical third (1 mm above the apical constriction). SkyScan 1172 desktop micro-CT system (Bruker), 13.61 μm isotropic voxel size.
Examiner training and reliability
All measurements were made by a single board-certified oral and maxillofacial radiologist experienced in micro-CT analysis of dental hard tissues. Before data collection, the examiner completed a standardised calibration session on pilot specimens, during which all morphometric parameters (canal length, mesiodistal (MD) and buccolingual (BL) canal diameters, minimum dentin thickness (MDT), greatest dentin thickness (GDT), isthmus identification, and accessory canal classification) were reviewed against published reference definitions. Intra-examiner reliability was assessed by re-analysing 20% of randomly selected specimens (n = 12) after a two-week wash-out interval, and agreement was excellent (intraclass correlation coefficient, two-way mixed-effects, absolute-agreement, single measure; ICC(A,1) = 0.95). The pilot specimens used for this calibration were independent teeth and were not included in the final study sample.
Number of roots
The number of roots per tooth was recorded, and each root canal was identified by axial-plane navigation from the canal orifice to the apical constriction6.
Presence and level of the MB2 canal
Continuous axial observations were made along the mesiobuccal root from the cemento-enamel junction (CEJ) to the apex6. MB2 was considered present only when a separate orifice, divided from the MB1 orifice by an intervening band of dentin, was identified and its canal space could be traced as a continuous lumen for at least 2 mm apical to the orifice. Canal spaces without a complete dentin partition between the two orifices, and shorter or discontinuous canal spaces, were not counted as MB2 but were classified as lateral ramifications. The root level at which the MB2 orifice arose was recorded as the coronal, middle, or apical third. In every MB2-positive tooth the MB2 orifice lay palatal to MB1, so orifice direction relative to MB1 was not analysed further. When the mesiobuccal root contained two canals, the larger principal canal was designated MB1 and the smaller additional canal MB2, following the established definition of MB2 as the second (accessory) canal of the mesiobuccal root13. This complete-dentin-partition criterion is more conservative than the definitions used in some earlier micro-CT studies, which also counted incompletely separated or shorter canal spaces as MB2; the stricter criterion yields a lower but more specific prevalence estimate and should be considered when comparing our findings with those reports (Fig. 3).
Representative micro-CT axial cross-sections of the MB2 canal, which lay palatal to the mesiobuccal (MB1) canal whenever present. (a) Tooth 42, coronal third: MB2 present. (b, c) Tooth 23: no MB2 at the coronal third (b), but MB2 present at the middle third (c).
Inter-orifice distances
Canal orifices were defined as the openings of the canal system at the base of the pulp chamber, standardised at 2 mm apical to the CEJ on axial slices. Horizontal distances were measured between the centre points of the MB1, MB2, distobuccal (DB), and palatal (P) canals. The pairs measured were MB1–MB2, MB1–DB, MB1–P, DB–P, MB2–P, and MB2–DB6 (Fig. 1b).
Canal length
Canal length was measured along the central axis of each canal between two reference levels: a fixed coronal reference standardised at 2 mm apical to the CEJ, and an apical reference. In specimens with an intact apex, the apical reference was the apical constriction (minor apical foramen ), identified on serial axial slices as the narrowest point of the canal before the lumen widened toward the apical foramen. In the few specimens with minimal physiological resorption (≤ 2 mm, the maximum permitted by the inclusion criteria) in which the apical constriction had been remodelled and could no longer be identified, the apical reference was instead taken as the most apical point of the continuous patent canal lumen (the resorbed apical terminus). Teeth with resorption exceeding 2 mm, or with internal or external pathological resorption, were excluded (see Selection criteria); consequently the apical reference never lay more than 2 mm coronal to the original anatomical apex.
Canal diameters at three root levels
Root thirds were defined as coronal (0–1/3 of root length from the CEJ), middle (1/3–2/3), and apical (2/3 to the apex). For each canal, three cross-sections were used to assess canal diameter and dentin thickness: a coronal section (1 mm below the orifice), an apical section (1 mm coronal to the apical constriction), and a mid-root section between them. In each section the canal centre was located, and the following were measured. The mesiodistal canal diameter (C1, M1, A1 for coronal, middle, and apical, respectively) was the distance (mm) of a line through the canal centre parallel to the mesiodistal plane, between the internal dentin walls. The buccolingual canal diameter (C2, M2, A2) was the corresponding distance parallel to the buccolingual plane, between the internal dentin walls. (Fig. 1d).
Canal cross-sectional shape
Canal cross-sectional shape was determined from the aspect ratio, defined as the ratio of the major to the minor canal diameter, following Jou et al.14. On each axial cross-section at the coronal, middle, and apical levels, the major diameter was measured as the maximum (caliper/Feret) diameter of the canal lumen and the minor diameter as the maximum width perpendicular to it. Each canal lumen was classified individually as round (ratio < 1.05), oval (1.05–2), long oval (> 2–4), or flattened (> 4). Because the aspect ratio expresses the larger relative to the smaller diameter, this classification is independent of canal orientation.
Dentin wall thickness
Dentin wall thickness was measured on the buccal, lingual, mesial, and distal aspects as the dimension (mm) of a line, parallel to the corresponding plane, delimited by the innermost and outermost dentin points whose projection passed through the canal centre. MDT was defined as the shortest distance (mm) from the canal centre to the external root surface, measured at the coronal, middle, and apical levels; its directional location was recorded as buccal (B), palatal (P), mesial (M), distal (D), mesiobuccal (MB), mesiolingual (ML), distobuccal (DB), or distopalatal (DP). The GDT was the largest such distance in the same cross-section. (Fig. 1c).
Isthmuses and accessory canals
An isthmus was defined as a narrow, ribbon-shaped communication between two canals within the same root that contains pulp or pulp-derived tissue15. Its presence, location (coronal, middle, or apical third), roof and floor levels, and length were recorded, together with the distance from the isthmus floor to the major apical foramen. The terms and definitions of the accessory canals followed the terminology of the of Ahmed et al. classification9. An accessory canal was defined as a small canal leaving the main root canal that communicates with the external root surface or the furcation, at any level (coronal, middle, or apical third)10, and an accessory foramen as the opening of such a canal on the external root surface.
Statistical analysis
Statistical analyses were performed in IBM SPSS Statistics v26.0 (IBM Corp., Armonk, NY, USA), with the significance level set at 0.05 (α = 0.05). Normality of continuous variables was assessed with the Shapiro–Wilk test and homogeneity of variance with Levene’s test. Inter-orifice distances were compared by one-way ANOVA with Tukey’s HSD post-hoc test, which provides strong control of the family-wise error rate. For canal diameters, 21 of 24 variables (87.5%) deviated significantly from normality, particularly at the middle and apical levels; the non-parametric Friedman test was therefore used for within-root comparisons across levels (with Bonferroni-corrected Wilcoxon signed-rank post-hoc tests), and Bonferroni-corrected Mann–Whitney U tests were used for pairwise MD versus BL comparisons within each root (family size = 3 per root; adjusted α = 0.0167). The distribution of canal cross-sectional shape across the three levels was analysed with the Fisher–Freeman–Halton exact test, owing to sparse expected cell counts, with the asymptotic Pearson χ² and Cramér’s V reported as supporting effect-size measures. To test for interactions between root level and measurement direction, a two-way repeated-measures ANOVA (Greenhouse–Geisser correction where appropriate) was performed separately for each root, with Level (coronal, middle, apical) and Direction (MD, BL) as within-subject factors; listwise deletion was applied, so the effective per-root n equals the number of specimens with complete MD and BL measurements at all three levels. DB and palatal canal diameters were measured in all 60 specimens, including the 10 two-rooted teeth with DB–P fusion, in which both canals remain individually traceable on micro-CT within the fused root mass; the reduced per-cell n reflects measurement losses (calcification, canal merging, indistinct margins) rather than exclusion of two-rooted specimens. Where one diameter direction could not be measured reliably at a given cross-section, the observation was retained for the available direction only, producing minor MD/BL n discrepancies within the same level. MDT and GDT were each recorded as a single global value per tooth: MDT as the shortest canal-wall-to-surface distance across the whole tooth, and GDT as the largest such distance, with the root and level of each also recorded. Because MDT concentrated at the apical level (71% of specimens) and GDT at the coronal level in the palatal root (79% of specimens), the cell sizes in the dentin-thickness summary (tabulated by root and level) are inherently unequal and reflect the true anatomical distribution rather than data-collection failure. Cells with n ≤ 3 are reported descriptively only, with no inferential testing.
Results
Of the 60 maxillary primary molars analysed, 50 (83.3%) had three roots and 10 (16.7%) had two roots, the latter representing the typical distobuccal–palatal fusion variant. An MB2 canal was identified in 26 mesiobuccal roots (43.3%), and in every MB2-positive tooth the orifice lay palatal to MB1. The MB2 orifice arose in the coronal third in 20 teeth (76.9% of MB2-positive teeth; 33.3% of the whole sample) and in the middle third in 6 teeth (23.1% of MB2-positive teeth; 10.0% of the whole sample); no MB2 orifice was found in the apical third (Table 1; Fig. 3). Across the 174 evaluable axial cross-sections, 315 canal lumens were classified by aspect ratio. Long-oval (152; 48.3%) and oval (143; 45.4%) profiles predominated, followed by flattened canals (20; 6.3%); no canal met the round criterion (< 1.05) (Fig. 4; Table 2). Canal shape did not differ significantly among root levels (Fisher–Freeman–Halton exact p = 0.15; asymptotic χ²(4) = 6.67; Cramér’s V = 0.10): although oval profiles were proportionally more frequent apically (56.5%) than coronally (44.5%), and long-oval profiles correspondingly less frequent (37.1% versus 51.6%), these differences did not reach significance (Fig. 5; Table 2). Wilson 95% confidence intervals (CIs) for the principal prevalence estimates were: MB2 43.3% (95% CI 31.6–55.9%); two-rooted teeth 16.7% (95% CI 9.3–28.0%).
Representative micro-CT axial cross-sections illustrating the canal shape categories used in this study, classified by the aspect ratio (major/minor canal diameter) following Jou et al.14: round (< 1.05), oval (1.05–2), long oval (> 2–4), and flattened (> 4). In each panel the major (green, maximum caliper) and minor (blue, perpendicular) canal diameters are overlaid, and the measured aspect ratio is given. Images are taken from specimens within the study sample. No canal in the sample met the round criterion (< 1.05), so the round category is not illustrated.
Canal cross-sectional shape distribution across the coronal, middle, and apical root levels (100% stacked bars; segment labels give the within-level percentage and count). Shape was classified from the major-to-minor caliper diameter ratio (round < 1.05, oval 1.05–2, long oval > 2–4, flattened > 4). No canal met the round criterion. The distribution did not differ significantly across levels (Fisher–Freeman–Halton exact p = 0.15; asymptotic χ²(4) = 6.67; Cramér’s V = 0.10).
Canal length
Canal length varied significantly across canal types (p < 0.001). Mean length was 8.59 ± 1.57 mm for MB1 (range 5.30–11.97 mm; n = 60), 8.40 ± 2.06 mm for the palatal canal (range 5.01–14.10 mm; n = 59), 7.29 ± 1.97 mm for the distobuccal canal (range 3.80–12.81 mm; n = 58), and 7.52 ± 2.44 mm for MB2 where measurable (range 2.88–12.42 mm; n = 20). Although MB2 was on average shorter than MB1, the difference was not significant after correction for multiple comparisons (Dunn’s test, p = 0.20); the significant differences were between MB1 and DB, and between the palatal and DB canals (Table 3).
Inter-orifice distances
Inter-orifice distances differed significantly across canal pairs (one-way ANOVA, F(5, 241) = 36.68, p < 0.001). The shortest mean separation was between MB1 and MB2 (1.12 ± 0.49 mm) and the greatest between MB1 and the palatal canal (5.54 ± 1.70 mm) (Table 4).
Canal diameters at coronal, middle, and apical levels
Buccolingual diameters significantly exceeded mesiodistal diameters in the MB1, MB2, and DB roots at all three levels (p ≤ 0.007), whereas the palatal root showed no significant directional difference at any level (p > 0.05) (Table 5).
Dentin wall thickness
Apical MDT was lowest in the palatal roots (0.10 ± 0.04 mm), followed by the mesiobuccal (0.12 ± 0.05 mm) and distobuccal (0.21 ± 0.19 mm) roots, a difference that was statistically significant across roots (Kruskal–Wallis p = 0.014). No significant inter-root differences were found at the middle or coronal levels (p > 0.05). GDT consistently exceeded MDT at all levels and in all roots (Table 6). Several cells in Table 6 contain small subsamples (n ≤ 3); these values are presented for descriptive purposes only, and the associated p-values should be interpreted with caution.
Isthmuses
Isthmuses were identified in the coronal third in 17 of 60 specimens (28.3%), in the middle third in 14 of 60 (23.3%), and in the apical third in only 4 of 60 (6.7%) (Fig. 6). The distobuccal–palatal (DB–P) region was the most frequent isthmus site at the coronal (17/60; 28.3%), middle (12/60; 20.0%), and apical (4/60; 6.7%) levels, whereas isthmuses between MB1 and MB2 were rare and, after the corrected MB2 classification, were confined to the middle third (2/60; 3.3%), with none in the coronal or apical third; representative MB1–MB2 isthmus cross-sections are shown in Fig. 7. No specimen showed a continuous isthmus from the coronal to the apical third. Wilson 95% CIs were: coronal 28.3% (18.5–40.8%), middle 23.3% (14.4–35.4%), and apical 6.7% (2.6–15.9%).
Isthmus presence and absence at the coronal, middle, and apical levels among the major canal pairs in 60 maxillary primary molars. Total isthmus prevalence: coronal 28.3% (DB–P 28.3% + MB1–MB2 0%); middle 23.3% (DB–P 20.0% + MB1–MB2 3.3%); apical 6.7% (DB–P 6.7% + MB1–MB2 0%).
Representative micro-CT cross-sections showing MB1–MB2 isthmuses (Specimens 35, 40, 24, and 25). In each panel the two mesiobuccal canal lumens are shaded cyan and the narrow ribbon-shaped isthmus connecting them is highlighted in red (arrow); canal lumens and the connecting isthmus were separated by local canal width (distance transform). MB1–MB2 isthmuses occurred predominantly in the coronal and middle thirds.
Accessory canals
Accessory canals were observed in all four canal types. They were most frequent in the distobuccal root (26/60; 43.3%), followed by MB1 (22/60; 36.7%), palatal (20/60; 33.3%), and MB2 (6/60; 10.0%). The level-by-direction distribution differed significantly within the distobuccal root (p = 0.001) but not in the MB1, MB2, or palatal roots (p = 0.097, 0.054, and 0.951, respectively) (Fig. 8). Accessory canals were most often located in the apical third and opened onto the external root surface; those opening into the furcation arose in the coronal furcation region, including within the fused distobuccal–palatal roots of the two-rooted teeth. Wilson 95% CIs by root were: distobuccal 43.3% (31.6–55.9%), MB1 36.7% (25.6–49.3%), palatal 33.3% (22.7–45.9%), and MB2 10.0% (4.7–20.1%) (Fig. 9).
Distribution of accessory canals according to root (MB1 = mesiobuccal canal 1; MB2 = second mesiobuccal canal; DB = distobuccal; P = palatal), location (B = buccal; P = palatal; M = mesial; D = distal), and root level (%). Y-axis = percentage of teeth (n = 60).
Representative micro-CT coronal section of a maxillary primary molar showing accessory (lateral) canals (white arrows) branching from the main root canal toward the external root surface: one in the cervical (coronal) third and one in the apical region. SkyScan 1172 desktop micro-CT system (Bruker), 13.61 μm isotropic voxel size; scale bar shown.
Discussion
This micro-CT study provides a detailed three-dimensional description of the root canal anatomy of maxillary primary molars. The main findings were: (i) MB2 canals were present in 43.3% of teeth, in a location palatal to MB1, and present mainly in the coronal third (20/26) rather than the middle third (6/26); (ii) the MB1, MB2, and DB roots showed a marked buccolingual-to-mesiodistal asymmetry, whereas the palatal root was close to circular; (iii) in the MB1, MB2, and DB roots, the buccolingual-to-mesiodistal ratio narrowed significantly from the coronal to the apical third (a level-by-direction interaction in the two-way RM ANOVA), so these canals became progressively less oval toward the apex; and (iv) the apical minimum dentin thickness was very small in the palatal (0.10 ± 0.04 mm) and mesiobuccal (0.12 ± 0.05 mm) roots. To the best of knowledge, this is among the first micro-CT studies to characterise all of these features together in a single sample of maxillary primary molars in this Egyptian cohort.
Three-rooted forms dominated the sample (83.3%), in keeping with the trifurcated anatomy that distinguishes maxillary primary molars from their mandibular counterparts2,16,17. The 16.7% two-rooted variant identified in this study is less common in some earlier work. For instance, Teixeira et al16. described all maxillary molar specimens in their study (n = 30) as three-rooted. The MB2 prevalence of 43.3% falls within the previously reported range for primary molars (20–60%)1,3 and is lower than the values reported for permanent maxillary first molars, where CBCT studies describe prevalences of roughly 60–70%, with a markedly lower prevalence (about 25–40%) in permanent maxillary second molars18,19,20. The relatively strict criterion applied here, counting MB2 as present only when a complete band of dentin separated its orifice from that of MB1, probably contributes to the moderate prevalence relative to studies that also count incompletely separated canal spaces.
Because the MB2 orifice is located palatal to MB1 all teeth included in this study, this consistent palatal position can guide the search for the orifice during access cavity preparation. Conventional two-dimensional radiography is known to miss a substantial proportion of canal complexities in permanent molars21, and cone beam computed tomography is not used clinically for primary molars because of radiation considerations and the limited remaining lifespan of these teeth. Case-specific volumetric assessment is therefore not a realistic clinical option, and the value of the present ex vivo characterisation lies in supplying the prior anatomical knowledge needed to anticipate features such as MB2 during conventional treatment, rather than in advocating volumetric imaging in the clinic.
Long-oval and oval canals together accounted for almost all cross-section (48.3% and 45.4%, respectively); flattened canals were uncommon (6.3%), and no canal met the strict round criterion (< 1.05). Canal shape did not vary significantly with root level (Fisher–Freeman–Halton exact p = 0.15), although oval profiles became proportionally more frequent toward the apex. This is consistent with earlier micro-CT and CBCT studies of primary molars reporting roundness indices of 0.26–0.63 across the MB, DB, and palatal roots2, and it resembles the pattern in permanent maxillary molars, in which oval or ribbon shapes predominate depending on root level22. An oval cross-section poses a dual clinical problem: oval canals cannot be fully shaped by rotary nickel-titanium (NiTi) instruments, which are round in cross-section and therefore leave the buccal and lingual extensions uninstrumented, raising the risk of biofilm persistence; conversely, larger or oval-path files risk strip perforation in roots with thin dentin walls. Recent evidence indicates that reciprocating single-file systems remove significantly more coronal dentin than rotary systems during primary molar pulpectomy23, which is of particular concern given the thin dentin walls documented here. Kinematics is not the only factor, however: instrument taper and apical size also affect the volume of dentin removed, with more aggressive taper and apical preparation tending to leave thinner residual walls24. When dentin conservation is the priority, conservative taper and tip selection should therefore take precedence over the rotary-versus-reciprocating distinction, and oval-specific techniques such as brushing motions, balanced-force filing, and activated irrigation are advisable regardless of the system used.
Inter-orifice distances differed significantly across canal pairs (p < 0.001). The shortest separation was between MB1 and MB2 (1.12 ± 0.49 mm), and this proximity substantially increases the risk of overlooking MB2 during access cavity preparation, given its 43.3% prevalence. The greatest separation was between MB1 and the palatal canal (5.54 ± 1.70 mm), with DB–P and MB1–DB forming an intermediate cluster. These distances are similar to those reported in earlier micro-CT studies of primary maxillary molars2 and comparable to permanent maxillary molars6, which reinforces the need for careful bur angulation and thorough axial navigation to reach every canal orifice, particularly MB2.
Within the MB1, MB2, and DB roots, the buccolingual dimension consistently and significantly exceeded the mesiodistal dimension at all three levels (p ≤ 0.007), and this asymmetry was most pronounced in the DB and MB1 roots. The palatal root, by contrast, showed no significant directional difference at any level (Mann–Whitney p > 0.05; two-way RM ANOVA Direction p= 0.399), indicating a more circular cross-section2.. This has direct clinical relevance: whereas oval MB and DB canals leave buccal and lingual recesses uninstrumented by round rotary files, the near-circular palatal canal can be shaped more completely with such files, without leaving large uninstrumented extensions25. Within-root taper from coronal to apical was substantial in both directions, which underlines the need for cautious apical shaping in the narrow mesiodistal dimension. The two-way RM ANOVA showed a significant level-by-direction interaction in the MB1 (p = 0.018) and DB (p = 0.001) roots, meaning that the buccolingual-to-mesiodistal ratio narrows from the coronal to the apical third and these canals become proportionally less oval toward the apex; in MB2 the buccolingual dimension also exceeded the mesiodistal at every level, but the level-by-direction interaction was not significant (p = 0.25). In the present sample, however, a discrete apical constriction was not identifiable in every canal. Where physiological resorption had altered the apical anatomy, the resorbed terminus was used as the apical reference (see Methods). Accordingly, the apical data reported here relate to the apical third rather than to a consistently defined apical constriction.
Apical MDT was thinnest in the palatal roots (0.10 ± 0.04 mm), followed by the mesiobuccal (0.12 ± 0.05 mm) and distobuccal (0.21 ± 0.19 mm) roots, and GDT consistently exceeded MDT at all levels. This asymmetry reflects the elliptical cross-section of primary molar root canals, in which the thinner wall is the true anatomical danger zone for perforation during instrumentation. The thin apical walls in the MB and P roots confirm a potential risk of strip perforation, in line with the thin-walled apical danger zone reported on another CBCT study in primary molars26, and they underline the need for conservative apical preparation to preserve residual dentin. The greater apical MDT in the DB root provides a comparatively safer margin for shaping compared to other roots.
Isthmuses showed a clear level-dependent gradient: 28.3% coronally, 23.3% in the middle third, and only 6.7% apically. The DB–P region was the most frequent isthmus location, whereas MB1–MB2 isthmuses were rare and, after the corrected MB2 classification, were confined to the middle third (3.3%). These prevalences are higher than those reported in some earlier micro-CT studies of primary maxillary molars2, and reflect the higher sensitivity of the present 13.61-µm protocol. They are substantially lower than the 69–80% MB-root isthmus prevalence reported for permanent maxillary molars6; a difference that may reflect thinner dentinal partitions and different developmental anatomy in primary teeth. Clinically, the high coronal and middle-third isthmus prevalence, particularly in the DB–P region and at the MB1–MB2 connection, is a debridement challenge that supports the use of activated irrigation and supplementary cleaning during pulpectomy.
Accessory canals were identified in all four canal types, which were most often in the DB root (43.3%) and least often in MB2 (10.0%). They were most commonly located in the apical third, consistent with previously reported patterns9. Pulpectomy success in primary teeth is multifactorial, and reported failure rates can be substantial; Dou et al.27 reported an approximately 37% survival rate at 30 months after pulpectomy of primary teeth in a single retrospective cohort. The anatomical complexity including the high prevalence of MB2, the oval cross-sections, and the DB–P isthmuses, is likely to contribute to such unfavorable outcomes when it is not adequately addressed during treatment.
In practical terms, these findings translate into several recommendations for pulpectomy of maxillary primary molars. Magnification (dental loupes or the dental operating microscope) may be helpful for MB2 detection, given its 43.3% prevalence and its mean 1.12 mm proximity to MB1. This characteristic anatomy suggests that circumferential instrumentation may be beneficial for the MB1, MB2, and DB roots, whereas conventional round instrumentation is generally adequate for the near-circular palatal canal. Apical preparation should be conservative, particularly in the palatal and mesiobuccal roots, where apical MDT approaches 0.1 mm. These clinical suggestions are inferred from the anatomical findings and were not tested clinically in the present study; they should be regarded as anatomically informed considerations rather than validated clinical recommendations.
This study has several limitations. First, its in vitro design limits direct extrapolation to clinical pulpectomy outcomes. Second, although 60 teeth provided adequate power for estimating the prevalence of common features, rare configurations (such as four-rooted molars or atypical isthmus patterns) may be under-represented. Finally, micro-CT findings cannot be transferred directly to clinical two-dimensional radiography, and the magnification advantage shown here is not routinely available in paediatric dental practice. In addition, the specimens were collected from two universities in Egypt and therefore do not represent the whole Egyptian population. The sample also comprised 41 primary first molars and 19 s molars. Because these subsamples were markedly unequal in size as a consequence of the study-specific inclusion criteria, a robust tooth-type-specific comparison between first and second molars was not feasible. Future work should correlate these three-dimensional anatomical findings with clinical pulpectomy outcomes and validate the MB2 diagnostic criteria across multiple operators and imaging systems.
Conclusion
Within the limitations of this in vitro study, maxillary primary molars showed complex and variable internal root canal anatomy. An MB2 canal was present in 43.3% of MB roots and, when present, it is located palatal to the MB1, mainly in the coronal third. The canals with oval and long-oval cross-sections predominated, without significant variation by root level. The MB and P roots shown considerable thinning of the apical dentin walls. Accessory canals were also frequent, most often in the distobuccal root and in the apical third, and represent an additional debridement challenge during pulpectomy. These three-dimensional features should be considered when planning access cavities, selecting instrumentation, and performing apical preparation during pulpectomy of maxillary primary molars.
Data availability
The datasets generated and analysed during the current study are available from the corresponding author on reasonable request.
References
Ramakrishnan, M., Niveditha, M. S. & Gurunathan, D. A short review on the root canal configuration of primary maxillary molars. Bioinformation 16, 1033–1036 (2020).
Fumes, A. C. et al. Root canal morphology of primary molars: A micro-computed tomography study. Eur. Arch. Paediatr. Dent. 15, 317–326 (2014).
Datta, P., Zahir, S., Kundu, G. & Dutta, K. An in vitro study of root canal system of human primary molars by using multidetector computed tomography. J. Indian Soc. Pedod. Prev. Dent. 37, 120–126 (2019).
Ahmed, H. M. A., Musale, P. K., Shahawy, E., Dummer, P. M. H. & O. I. & Application of a new system for classifying tooth, root and canal morphology in the primary dentition. Int. Endod J. 53, 27–35 (2020).
Rouhani, A. et al. Prevalence of missed canals in endodontically treated teeth: a cone-beam computed tomography study. J. Clin. Exp. Dent. 15, e605–e611 (2023).
Tonelli, S. Q. et al. Micro-CT evaluation of maxillary first molars: interorifice distances and internal anatomy of the mesiobuccal root. Braz Oral Res. 35, e060 (2021).
Ahmed, H. M. A., Versiani, M. A., De-Deus, G. & Dummer, P. M. H. A new system for classifying root and root canal morphology. Int. Endod. J. 50, 761–770 (2017).
Ahmed, H. M. A., Khamis, M. F. & Gutmann, J. L. Seven root canals in a deciduous maxillary molar detected by the dental operating microscope and micro-computed tomography. Scanning 38, 554–557 (2016).
Ahmed, H. M. A., Neelakantan, P. & Dummer, P. M. H. A new system for classifying accessory canal morphology. Int. Endod. J. 51, 164–176 (2018).
De Deus, Q. D. Frequency, location, and direction of the lateral, secondary, and accessory canals. J. Endod. 1, 361–366 (1975).
Keleş, A. & Keskin, C. A micro-computed tomographic study of band-shaped root canal isthmuses, having their floor in the apical third of mesial roots of mandibular first molars. Int. Endod J. 51, 240–246 (2018).
Siqueira, J. F. Jr. & Rôças, I. N. Present status and future directions: microbiology of endodontic infections. Int. Endod J. 55 (Suppl. 3), 512–530 (2022).
Kulild, J. C. & Peters, D. D. Incidence and configuration of canal systems in the mesiobuccal root of maxillary first and second molars. J. Endod. 16, 311–317 (1990).
Jou, Y. T., Karabucak, B., Levin, J. & Liu, D. Endodontic working width: current concepts and techniques. Dent. Clin. North. Am. 48, 323–335 (2004).
Hsu, Y. Y. & Kim, S. The resected root surface: the issue of canal isthmuses. Dent. Clin. North. Am. 41, 529–540 (1997).
Teixeira, T. F. et al. Deciduous molars complexity anatomy revealed by computed microtomography. Eur. J. Dent. 18, 789–795 (2024).
Ziya, M., Yüksel, B. N. & Sarı, Ş. Root canal morphology of mandibular primary molars: a micro-CT study. Cumhuriyet Dent. J. 22, 382–389 (2019).
Magat, G. & Hakbilen, S. Prevalence of second canal in the mesiobuccal root of permanent maxillary molars from a Turkish subpopulation: a cone-beam computed tomography study. Folia Morphol. (Warsz). 78, 351–358 (2019).
Martins, J. N. R. & Versiani, M. A. CBCT and micro-CT on the study of root canal anatomy. In The Root Canal Anatomy in Permanent Dentition (eds Versiani, M. A. et al.) 89–180 (Springer International Publishing, 2019).
Onn, H. Y., Sikun, M. S. Y. A., Rahman, A., Dhaliwal, J. S. & H. & Prevalence of mesiobuccal-2 canals in maxillary first and second molars among the Bruneian population: CBCT analysis. BDJ Open. 8, 32 (2022).
Vizzotto, M. B. et al. CBCT for the assessment of second mesiobuccal (MB2) canals in maxillary molar teeth: effect of voxel size and presence of root filling. Int. Endod J. 46, 870–876 (2013).
Chang, J. W. W. et al. Morphotypes of the apical constriction of maxillary molars: a micro-computed tomographic evaluation. Restor. Dent. Endod. 47, e19 (2022).
Eren, İ. & Sezer, B. Comparative evaluation of the remaining dentin volume following instrumentation with rotary, reciprocating, and hand files during root canal treatment in primary molars: an ex vivo study. J. Dent. Sci. 19, 2126–2134 (2024).
Seema, T., Ahammed, H., Parul, S. & Cheranjeevi, J. Comparative evaluation of dentin removal and taper of root canal preparation of Hand K file, ProTaper rotary file, and Kedo-S rotary file in primary molars using cone-beam computed tomography. Int. J. Clin. Pediatr. Dent. 13, 332–336 (2020).
Paqué, F., Balmer, M., Attin, T. & Peters, O. A. Preparation of oval-shaped root canals in mandibular molars using nickel-titanium rotary instruments: a micro-computed tomography study. J. Endod. 36, 703–707 (2010).
Justiniano-Navarro, C., Caballero-García, S., López-Rodriguez, G. & Geller-Palti, D. Evaluating the thickness of the root canal dentin wall in primary first molars using cone-beam computed tomography. Int. J. Clin. Pediatr. Dent. 16 (Suppl. 2), S122–S127 (2023).
Dou, G. et al. A retrospective study on the long-term outcomes of pulpectomy and influencing factors in primary teeth. J. Dent. Sci. 17, 771–779 (2022).
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. The authors received no financial support for the research, authorship, or publication of this article.
Author information
Authors and Affiliations
Department of Pedodontics and Oral Health, Faculty of Dental Medicine for Girls, Al- Azhar University, Cairo, Egypt
Alaa M. Eldehna & Nouralhouda F. Abd-Allah
Department of Endodontics, Faculty of Dentistry, Ain Shams University, Cairo, Egypt
Sherief Elzahar
Private Clinic, İstanbul, Turkey
Sherief Elzahar
Department of Oral and Maxillofacial Radiology, Faculty of Dentistry, October 6 University, Cairo, Egypt
Ahmad H. Ahmad
Pediatric Dentistry and Dental Public Health Department, October 6 University, Cairo, Egypt
Dina Yousry El Kharadly, Nada Adel Salem & Dina M. Abdel-Ghany
Department of Oral and Dental Biology, Faculty of Dental Medicine for Girls, Al-Azhar University, Cairo, Egypt
Mena F. Abd-Allah
Department of Restorative Dentistry, Faculty of Dentistry, Universiti Malaya, Kuala Lumpur, Malaysia
Hany Mohamed Aly Ahmed
Department of Endodontics, Faculty of Dental Medicine for Girls, Al-Azhar University, Cairo, Egypt
Heba Abdelkafy
Department of Endodontics, , Faculty of Oral and Dental Surgery,Misr University of Science and Technology , Giza, Egypt
Heba Abdelkafy
Authors
- Alaa M. Eldehna
- Sherief Elzahar
- Ahmad H. Ahmad
- Dina Yousry El Kharadly
- Nada Adel Salem
- Dina M. Abdel-Ghany
- Nouralhouda F. Abd-Allah
- Mena F. Abd-Allah
- Hany Mohamed Aly Ahmed
- Heba Abdelkafy
Contributions
A.M.E. Conceptualization, Methodology, Formal analysis and investigation, Visualization, Writing—original draft, and Writing—review and editing. S.E. Conceptualization, Methodology, Formal analysis and investigation, and conducting the research and investigation process. A.H.A. Formal analysis and investigation, Data curation, data collection, and conducting the research and investigation process. D.Y.E.K. Resources, Conceptualization, Data collection, provision of study materials, and Data curation. N.A.S. Project administration, Resources, and Supervision. D.M.A. Resources, provision of study materials, and Supervision. N.F.A. Resources, Conceptualization, Data collection, and conducting the research and investigation process. M.F.A. Formal analysis and investigation, Data collection, and conducting the research and investigation process. H.An. Visualization, Writing—review and editing, and conducting the research and investigation process. H.Ab. Writing—review and editing, Methodology, Formal analysis and investigation, Data collection, conducting the research and investigation process, and Supervision.
Corresponding authors
Correspondence to Alaa M. Eldehna or Heba Abdelkafy.
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Eldehna, A.M., Elzahar, S., Ahmad, A.H. et al. Micro-computed tomography analysis of root and canal morphology in maxillary primary molars. Sci Rep 16, 33592 (2026). https://doi.org/10.1038/s41598-026-71739-0
Received: 19 July 2026
Accepted: 10 September 2026
Published: 23 October 2026
Version of record: 23 October 2026
DOI: https://doi.org/10.1038/s41598-026-71739-0
Keywords
来源:Nature 健康护理主题 · nature.com