Effect of positional errors on the accuracy of cervical vertebrae maturation assessment using CBCT and lateral cephalograms

Published:November 05, 2020DOI:https://doi.org/10.1016/j.ejwf.2020.09.006

      Abstract

      Background

      The objective of this study was to evaluate the effects of single plane and multiplane rotational errors in yaw, pitch, and roll of the head while recording the lateral cephalogram on CVM (cervical vertebrae maturity) assessment.

      Methods

      A total of 40 cone-beam computed tomography (CBCT) scans and 360 lateral cephalograms were analyzed for patients with different rotations: Controls (no rotation), Y5 (yaw 5° rotation), Y10 (yaw 10° rotation), R5 (roll 5° rotation), R10 (Roll 10° rotation), P5 (pitch 5° rotation), P10 (pitch 10° rotation), YRP5 (yaw, roll, and pitch 5° rotation), and YRP10 (yaw, roll, and pitch 10° rotation). The C2, C3, and C4 concavity and their base-anterior ratio and posterior-anterior ratio were measured. In addition, maxillomandibular linear parameters, such as effective mandibular length and height, mandibular body length, effective midface length, and maxillomandibular differential, were also evaluated.

      Results

      Y5, Y10, R5, and R10 led to overestimation of CVM in comparison with controls. Multiplane rotations (YRP5 and YRP10) led to more inaccuracies in CVM measurements than single plane rotations; 10° of rotation led to more inaccuracies than 5° of rotation while recording the lateral cephalogram, irrespective of the plane. Yaw rotational errors led to an underestimation of maxillomandibular linear measurements, whereas roll rotational errors led to an overestimation of the measurements; however, there were wide individual variations in the measurements between the different rotations and controls.

      Conclusions

      Rotational errors lead to overestimation of CVM assessment. Multiplane rotations cause higher inaccuracies than single plane rotations. Increased degree of rotations while capturing the lateral cephalograms lead to more inaccuracies in CVM assessment.

      Keywords

      1. Introduction

      Identification of the skeletal maturity of a patient is important when planning growth modification or orthopedic correction. Different indices have been proposed to identify skeletal maturation [
      • Lewis A.B.
      • Garn S.M.
      The relationship between tooth formation and other maturation factors.
      ,
      • Hunter C.J.
      The correlation of facial growth with body height and skeletal maturation at adolescence.
      ,
      • Fishman L.S.
      Maturational patterns and prediction during adolescence.
      ]. Of these, the cervical vertebrae maturation (CVM) method has been used by orthodontists frequently, as it can be performed using the lateral cephalogram, which is routinely obtained for orthodontic diagnosis. Thus, CVM has emerged as a popular tool for assessment of the growth status of the patient. Studies have demonstrated the utility of CVM in diagnosis and treatment timing of patients who may require functional orthopedic appliances and orthognathic surgery [
      • Franchi L.
      • Baccetti T.
      New emphasis on the role of mandibular skeletal maturity in dentofacial orthopedics.
      ,
      • Baccetti T.
      • Franchi L.
      • Toth L.R.
      • McNamara Jr., J.A.
      Treatment timing for Twin-block therapy.
      ].
      The correlation of CVM with the growth of mandible has been studied with conflicting results [
      • O'Reilly M.T.
      • Yanniello G.J.
      Mandibular growth changes and maturation of cervical vertebrae--a longitudinal cephalometric study.
      ,
      • Franchi L.
      • Baccetti T.
      • McNamara Jr., J.A.
      Mandibular growth as related to cervical vertebral maturation and body height.
      ]. A few studies have reported that CVM is effective in analyzing the growth of the mandible, whereas other studies did not show any positive correlation between CVM and mandibular growth [
      • Beit P.
      • Peltomäki T.
      • Schätzle M.
      • Signorelli L.
      • Patcas R.
      Evaluating the agreement of skeletal age assessment based on hand-wrist and cervical vertebrae radiography.
      ,
      • Ball G.
      • Woodside D.
      • Tompson B.
      • Hunter W.S.
      • Posluns J.
      Relationship between cervical vertebral maturation and mandibular growth.
      ]. It has been reported that changes in the head position may lead to errors in cephalometric analysis, which is likely to have an effect on the accuracy of CVM measurements for the assessment of the growth of the mandible [
      • Major P.W.
      • Johnson D.E.
      • Hesse K.L.
      • Glover K.E.
      Effect of head orientation on posterior anterior cephalometric landmark identification.
      ]. Therefore, analysis of the effects of the head positioning errors on CVM measurements is important in that it may have a huge clinical impact on diagnosis and treatment planning.
      Lateral cephalogram gives a two-dimensional (2D) representation of three-dimensional (3D) structures in the head and neck, and this 2D representation may lead to inaccuracies in CVM assessment. Three-dimensional radiographs, such as cone-beam computed tomography (CBCT), have enabled orthodontists to overcome the limitations of 2D radiographs; thus, more emphasis has been placed recently on investigating the utility of CBCTs in orthodontics. The reliability of CBCT for the assessment of CVM compared with lateral cephalogram have been reported with conflicting results [
      • Shim J.J.
      • Heo G.
      • Lagravère M.O.
      Assessment of skeletal maturation based on cervical vertebrae in CBCT.
      ,
      • Joshi V.
      • Yamaguchi T.
      • Matsuda Y.
      • Kaneko N.
      • Maki K.
      • Okano T.
      Skeletal maturity assessment with the use of cone-beam computerized tomography.
      ]. Such conflicting results warrant further investigations. In addition, the literature on the accuracy of CVM assessment between CBCT and lateral cephalogram is scarce. Thus, in addition to identifying the effects of head rotation on CVM assessment, a comparison of CVM indicators between CBCT and lateral cephalogram was also performed in this study.
      The objectives of this study were as follows: 1) to determine whether there is any difference in the CVM indicators identified on the CBCT and the lateral cephalogram; 2) to evaluate the accuracy of CVM indicators on the lateral cephalogram with changes in yaw, pitch, and roll of the head position; 3) to analyze the effect of the degree of positional errors on CVM assessment; and 4) compare and contrast the effect of single plane versus multiplane positional errors on the accuracy of CVM evaluation. Our null hypothesis was that there is no difference in the assessment of CVM staging on lateral cephalograms with yaw, pitch, and roll rotation of the head.

      2. Materials and methods

      This study was undertaken at the University of Connecticut (the institutional review board approval number was SM 1168). The following were the inclusion criteria: patients aged 11 to 15 years with no history of prior orthodontics or temporomandibular joint disorder (TMD); absence of any systemic disorders; and presence of cervical vertebrae C2, C3, and C4 on the CBCT. CBCTs of 45 patients were initially evaluated, but 5 were excluded because of an inadequate field of view. This retrospective study reviewed 40 CBCT images of patients (mean age 13.66 ± 1.14 years) seeking orthodontic treatment and lateral cephalograms generated from those CBCT images. CBCT scans were acquired using the iCAT Next Generation (Imaging Sciences International, Hatfield, PA) CBCT unit. A standardized protocol for the iCAT with 0.25 mm voxel size, 8.9 seconds acquisition time with 120 kV, and 20 mA was used. All scans were analyzed using Dolphin software (Version 11.9; Patterson Dental, Chatsworth, CA). The orientation of the CBCTs was performed in a standardized manner (Supplemental Fig. 1) [
      • Kim H.J.
      • Hong M.
      • Park H.S.
      Analysis of dental compensation in patients with facial asymmetry using cone-beam computed tomography.
      ]. At this orientation, the lateral cephalograms were generated with perspective projection (with default settings of 9.7% magnification, Emitter to Patient distance of 1550 mm, 100 mm ruler) maintaining the Mid-plane to Film distance of 150 mm (Bolton-Broadbent dimensions) and defined as controls (lateral cephalogram at 0° rotation). The malocclusion of the patients evaluated in the study is shown in Supplemental Table 1a.
      Yaw was defined as rotation in the horizontal plane, roll as rotation in the frontal plane, and pitch as rotation in the sagittal plane [
      • Ackerman J.L.
      • Proffit W.R.
      • Sarver D.M.
      • Ackerman M.B.
      • Kean M.R.
      Pitch, roll, and yaw: describing the spatial orientation of dentofacial traits.
      ]. The orientation of the CBCTs was modified to generate yaw, pitch, and roll rotation of 5° and 10°. The lateral cephalograms generated from these CBCTs were defined as Y5 (yaw 5° rotation), Y10 (yaw 10° rotation; Fig. 1), R5 (roll 5° rotation), R10 (roll 10° rotation; Fig. 2), P5 (pitch 5° rotation), P10 (pitch 10° rotation; Fig. 3), YRP5 (yaw, roll, and pitch 5° rotation), and YRP10 (yaw, roll, and pitch 10° rotation; Fig. 4). Thus, there were nine different orientations performed for each CBCT, and a lateral cephalogram was generated from the CBCT for each rotation. Each CBCT led to the generation of nine lateral cephalograms; therefore, with 40 CBCTs, 40 lateral cephalograms were generated with each of the nine rotations, resulting in a total of 360 lateral cephalograms.
      Figure thumbnail gr1
      Fig. 1Orientation of CBCT and generation of lateral cephalograms with yaw rotation: (A) 0° rotation (ideal rotation, n = 40); (B) control lateral cephalogram (n = 40); (C) 5° yaw rotation; (D) Y5 lateral cephalogram (n = 40); (E) 10° yaw rotation; (F) Y10 lateral cephalogram (n = 40).
      Figure thumbnail gr2
      Fig. 2Orientation of CBCT and generation of lateral cephalograms with roll rotation: (A) side view with 5° roll rotation; (B) side view with 10° roll rotation; (C) frontal view with 5° roll rotation; (D) R5 lateral cephalogram (n = 40); (E) 10° roll rotation; (F) R10 lateral cephalogram (n = 40).
      Figure thumbnail gr3
      Fig. 3Orientation of CBCT and generation of lateral cephalograms with pitch rotations: (A) side view with 5° pitch rotation; (B) Side view with 10° pitch rotation; (C) frontal view showing orientation of CBCT with 5° pitch rotation; (D) P5 lateral cephalogram (n = 40); (E) orientation of CBCT with 10° pitch rotation; (F) P10 lateral cephalogram (n = 40).
      Figure thumbnail gr4
      Fig. 4Orientation of CBCT and generation of lateral cephalograms with multiplane rotations: (A) side with 5° YRP rotation; (B) side view with 10° YRP rotation; (C) frontal view with 5° YRP rotation; (D) YRP5 lateral cephalogram (n = 40); (E) Fontal view with 10° YRP rotation; (F) YRP10 lateral cephalogram (n = 40).
      The C2, C3, and C4 concavity and their base-anterior ratio (BAR) and posterior-anterior ratio (PAR) were measured [
      • Baccetti T.
      • Franchi L.
      • McNamara Jr., J.A.
      An improved version of the cervical vertebral maturation (CVM) method for the assessment of mandibular growth.
      ]. In addition, maxillomandibular linear parameters, such as effective mandibular length (ML) and height (MH), mandibular body length (MBL), effective midface length (MFL), and maxillomandibular differential (MMD) were also evaluated. Landmarks on the CBCT were identified using the coronal, axial, and sagittal slices (Supplemental Fig. 2). The landmarks and parameters used in the study are described in Supplemental Table 1b and Figure 5. For the bilateral landmarks, such as condylion and gonion, measurements were performed on the left and the right sides individually and the mean values were considered for analysis. All the measurements were performed by a single investigator (SM) under standardized conditions of ambient light and sound. Twenty random samples of CBCTs and lateral cephalograms were measured by the same investigator after a period of 4 weeks and by another investigator (RD) to evaluate the intrarater and interrater reliability.
      Figure thumbnail gr5
      Fig. 5(A) Diagrammatic representation of the landmarks used in the study; (B) Measurements used in the study.

      2.1 Statistical analysis

      The post hoc power analysis was performed by resampling data (n = 1000) using the R package ‘simr’ [
      • Green P.
      • Macleod C.J.
      SIMR: An R package for power analysis of generalized linear mixed models by simulation.
      ]. A sample of 40 would allow us to detect a 15% of change from 5 or 10-degree rotation compared with controls for more than 99% power at the 5% significance level, using a linear mixed-effects model with a random patient intercept.
      The mean difference between controls and CBCTs was compared using a two-sided paired t-test. Lateral cephalograms at different head rotations in pitch, roll, yaw, or their combinations plus controls were compared for continuous parameters using a linear mixed-effects model with a random patient intercept. Pairwise comparisons were conducted using the least square means method with P values adjusted for multiple testing by Tukey's method. The Fleiss kappa was calculated to assess the agreement among measurements, 0.41–0.60 as moderate, 0.61–0.80 as substantial, and 0.81–1.00 almost perfect. Intraclass correlation coefficients were calculated to assess the consistency for repeated measurements within and between the raters. All the statistical analyses were performed in R version 3.6.1.

      3. Results

      3.1 Comparison between CBCT and controls

      The C2 and C4 concavity, ML, MH, MBL, MFL, and MMD were significantly greater in controls as compared with the CBCT (P < 0.05, Table 1). A higher percentage of patients were classified as having deep C2, C3, and C4 concavity in controls compared with the CBCT group (Supplemental Table 1c). The Fleiss’ kappa coefficient showed moderate agreement for C2 and C4 concavity between CBCT and controls (Supplemental Table 1d).
      Table 1Difference between cone beam computed tomography (CBCT, n = 40) and lateral cephalogram generated at 0 rotation (Controls, n = 40)
      ParametersCBCTControlsCBCT-controlsP value
      Significant at P value ≤ 0.05.
      Mean (SD)Mean (SD)Mean (95% CI)
      C2 Concavity (mm)1.18 (0.69)1.42 (0.73)−0.24 (−0.36 to −0.12)0.000
      Significant at P value ≤ 0.05.
      C3 Concavity (mm)1.06 (0.55)1.21 (0.78)−0.15 (−0.33 to 0.03)0.096
      C4 Concavity (mm)0.75 (0.59)0.96 (0.65)−0.22 (−0.37 to −0.06)0.007
      Significant at P value ≤ 0.05.
      C3 Base-Anterior Ratio (%)110.14 (23.03)106.72 (27.39)3.43 (−1.43 to 8.28)0.161
      C3 Posterior-Anterior Ratio (%)119.13 (17.49)110.79 (18.12)8.34 (3.11 to 13.57)0.003
      Significant at P value ≤ 0.05.
      C4 Base-Anterior Ratio (%)115.19 (21.12)113.55 (22.27)1.65 (−4.67 to 7.97)0.601
      C4 Posterior-Anterior Ratio (%)118.03 (13.99)116.54 (16.52)1.49 (−3.73 to 6.71)0.567
      Effective Mandibular Length (Co-Gn) (mm)113.38 (7.4)120.63 (7.86)−7.26 (−8.21 to −6.3)0.000
      Significant at P value ≤ 0.05.
      Mandibular Height (Co-Go) (mm)61.03 (6.91)66.79 (7.13)−5.76 (−6.54 to −4.97)0.000
      Significant at P value ≤ 0.05.
      Mandibular Body Length (Go-Gn) (mm)60.53 (4.26)66.45 (4.55)−5.92 (−6.51 to −5.33)0.000
      Significant at P value ≤ 0.05.
      Effective Midface Length (Co-A) (mm)81.48 (4.74)86.61 (5.16)−5.13 (−5.84 to −4.42)0.000
      Significant at P value ≤ 0.05.
      Maxillomandibular Differential (Co-Gn - Co-A) (mm)31.89 (4.87)34.04 (4.79)−2.15 (−2.98 to −1.32)0.000
      Significant at P value ≤ 0.05.
      A, A point; CI, confidence interval; Co, condyle; Gn, gnathion; Go, gonion; SD, standard deviation.
      a Significant at P value ≤ 0.05.

      3.2 Comparison between controls, Y5, and Y10

      MFL and MMD were sensitive to yaw rotation of 5° and 10°. The C2, C3, and C4 concavity were not significantly different in the Y5 and controls. The C4 concavity, C4-BAR, and MMD were significantly increased (P < 0.05), and the C4-PAR, ML, MH, MBL, and MFL were significantly decreased (P < 0.05) in Y10 as compared with controls (Table 2). The percentage of patients with C2, C3, and C4 concavity classified as deep was higher in Y5 and Y10 compared with controls (Supplemental Table 2a). Moderate agreement was found between controls and Y5 and between controls and Y10 for C2 and C3 concavity (Supplemental Table 2b).
      Table 2Difference between lateral cephalogram generated at 0 rotation (Controls, n = 40) and yaw rotation of 5° (Y5, n = 40) and yaw rotation of 10° (Y10, n = 40)
      ParametersControlsY5Y10Overall P valueY5- controlsY10- controlsY10- Y5
      Mean (SD)Mean (SD)Mean (SD)P value
      Significant at P value ≤ 0.05.
      Mean (95% CI)P value
      Significant at P value ≤ 0.05.
      Mean (95% CI)P value
      Significant at P value ≤ 0.05.
      Mean (95% CI)P value
      Significant at P value ≤ 0.05.
      C2 Concavity (mm)1.42 (0.73)1.53 (0.72)1.5 (0.69)0.2370.11 (−0.03 to 0.24)0.2310.08 (−0.05 to 0.2)0.446−0.03 (−0.15 to 0.1)0.902
      C3 Concavity (mm)1.21 (0.78)1.23 (0.72)1.29 (0.72)0.5700.02 (−0.13 to 0.18)0.9510.08 (−0.09 to 0.25)0.5570.06 (−0.08 to 0.19)0.744
      C4 Concavity (mm)0.96 (0.65)1.06 (0.68)1.2 (0.58)0.005
      Significant at P value ≤ 0.05.
      0.1 (−0.06 to 0.26)0.3650.24 (0.11 to 0.37)0.003
      Significant at P value ≤ 0.05.
      0.14 (0 to 0.29)0.121
      C3 Base-Anterior Ratio (%)106.72 (27.39)106.8 (25.66)107.61 (24.74)0.9150.08 (−4.82 to 4.97)0.9990.89 (−3.4 to 5.17)0.9240.81 (−4.15 to 5.77)0.936
      C3 Posterior-Anterior Ratio (%)110.79 (18.12)112.5 (15.13)111.25 (15.8)0.6721.72 (−2.54 to 5.97)0.6650.47 (−3.9 to 4.83)0.970−1.25 (−4.6 to 2.1)0.804
      C4 Base-Anterior Ratio (%)113.55 (22.27)115.32 (21.81)120.86 (24.78)0.012
      Significant at P value ≤ 0.05.
      1.77 (−2.79 to 6.34)0.7577.32 (1.75 to 12.89)0.012
      Significant at P value ≤ 0.05.
      5.54 (0.63 to 10.46)0.073
      C4 Posterior-Anterior Ratio (%)116.54 (16.52)112.94 (16.27)109.22 (16.9)0.003
      Significant at P value ≤ 0.05.
      −3.59 (−7.81 to 0.62)0.190−7.32 (−11.87 to −2.77)0.002
      Significant at P value ≤ 0.05.
      −3.73 (−7.3 to −0.16)0.168
      Effective Mandibular Length (Co-Gn) (mm)120.63 (7.86)120.33 (8.25)117.38 (7.63)0.000
      Significant at P value ≤ 0.05.
      −0.3 (−1.02 to 0.41)0.692−3.25 (−4.03 to −2.48)0.000
      Significant at P value ≤ 0.05.
      −2.95 (−3.71 to −2.2)0.000
      Significant at P value ≤ 0.05.
      Mandibular Height (Co-Go) (mm)66.79 (7.13)66.55 (7.42)65.05 (6.78)0.000
      Significant at P value ≤ 0.05.
      −0.24 (−1.04 to 0.57)0.811−1.74 (−2.5 to −0.98)0.000
      Significant at P value ≤ 0.05.
      −1.5 (−2.27 to −0.74)0.001
      Significant at P value ≤ 0.05.
      Mandibular Body Length (Go-Gn) (mm)66.45 (4.55)66.36 (4.41)65.59 (4.62)0.005
      Significant at P value ≤ 0.05.
      −0.1 (−0.62 to 0.42)0.931−0.86 (−1.51 to −0.22)0.007
      Significant at P value ≤ 0.05.
      −0.76 (−1.28 to −0.25)0.020
      Significant at P value ≤ 0.05.
      Effective Midface Length (Co-A) (mm)86.61 (5.16)85.25 (5.89)82.05 (5.42)0.000
      Significant at P value ≤ 0.05.
      −1.37 (−2.02 to −0.71)0.001
      Significant at P value ≤ 0.05.
      −4.56 (−5.37 to −3.74)0.000
      Significant at P value ≤ 0.05.
      −3.19 (−3.92 to −2.47)0.000
      Significant at P value ≤ 0.05.
      Maxillomandibular Differential (Co-Gn - Co-A) (mm)34.02 (4.78)35.08 (4.95)35.32 (4.86)0.000
      Significant at P value ≤ 0.05.
      1.06 (0.45 to 1.68)0.004
      Significant at P value ≤ 0.05.
      1.3 (0.73 to 1.88)0.000
      Significant at P value ≤ 0.05.
      0.24 (−0.5 to 0.98)0.734
      A, A point; CI, confidence interval; Co, condyle; Gn, gnathion; Go, gonion; SD, standard deviation.
      a Significant at P value ≤ 0.05.

      3.3 Comparison between controls, R5, and R10

      R5 showed significantly increased (P < 0.05) ML, MH, and MFL than controls. R10, when compared with R5 and controls, showed significantly increased C2, C3, and C4 concavity, C3-BAR, C3-PAR, C4-BAR, C4-PAR, ML, and MBL (P < 0.05, Table 3). The percentage of patients with C2, C3, and C4 concavity classified as deep was higher in R10 compared with controls and R5 (Supplemental Table 3a). Supplemental Table 3b shows that C2 and C3 concavity exhibited moderate agreement between controls and R5 and between controls and R10 for C2 concavity.
      Table 3Difference between lateral cephalogram generated at 0 rotation (Controls, n = 40) and roll rotation of 5° (R5, n = 40) and roll rotation of 10° (R10, n = 40)
      ParametersControlsR5R10Overall P valueR5- controlsR10- controlsR10- R5
      Mean (SD)Mean (SD)Mean (SD)P value
      Significant at P value ≤ 0.05.
      Mean (95% CI)P value
      Significant at P value ≤ 0.05.
      Mean (95% CI)P value
      Significant at P value ≤ 0.05.
      Mean (95% CI)P value
      Significant at P value ≤ 0.05.
      C2 Concavity (mm)1.42 (0.73)1.44 (0.68)1.77 (0.76)0.000
      Significant at P value ≤ 0.05.
      0.02 (−0.12 to 0.16)0.9660.34 (0.19 to 0.5)0.000
      Significant at P value ≤ 0.05.
      0.33 (0.19 to 0.46)0.000
      Significant at P value ≤ 0.05.
      C3 Concavity (mm)1.21 (0.78)1.25 (0.73)1.54 (0.74)0.000
      Significant at P value ≤ 0.05.
      0.04 (−0.12 to 0.21)0.8560.33 (0.15 to 0.51)0.000
      Significant at P value ≤ 0.05.
      0.29 (0.15 to 0.43)0.001
      Significant at P value ≤ 0.05.
      C4 Concavity (mm)0.96 (0.65)1 (0.62)1.31 (0.57)0.000
      Significant at P value ≤ 0.05.
      0.04 (−0.12 to 0.2)0.8570.35 (0.2 to 0.5)0.000
      Significant at P value ≤ 0.05.
      0.31 (0.16 to 0.46)0.000
      Significant at P value ≤ 0.05.
      C3 Base-Anterior Ratio (%)106.72 (27.39)107.53 (21.7)114.64 (25.07)0.000
      Significant at P value ≤ 0.05.
      0.82 (−3.87 to 5.51)0.9207.93 (4.17 to 11.68)0.001
      Significant at P value ≤ 0.05.
      7.11 (2.85 to 11.37)0.003
      Significant at P value ≤ 0.05.
      C3 Posterior-Anterior Ratio (%)110.79 (18.12)111.59 (12.3)122.16 (13.93)0.000
      Significant at P value ≤ 0.05.
      0.8 (−3.84 to 5.43)0.94211.37 (5.47 to 17.27)0.000
      Significant at P value ≤ 0.05.
      10.57 (6.69 to 14.46)0.000
      Significant at P value ≤ 0.05.
      C4 Base-Anterior Ratio (%)113.55 (22.27)114.85 (21.48)120.52 (19.9)0.002
      Significant at P value ≤ 0.05.
      1.3 (−2.97 to 5.58)0.7916.98 (2.38 to 11.58)0.002
      Significant at P value ≤ 0.05.
      5.68 (2.61 to 8.74)0.015
      Significant at P value ≤ 0.05.
      C4 Posterior-Anterior Ratio (%)116.54 (16.52)115.22 (13.37)125.06 (15.5)0.001
      Significant at P value ≤ 0.05.
      −1.32 (−5.92 to 3.28)0.8848.52 (2.3 to 14.75)0.009
      Significant at P value ≤ 0.05.
      9.84 (3.88 to 15.81)0.002
      Significant at P value ≤ 0.05.
      Effective Mandibular Length (Co-Gn) (mm)120.63 (7.86)122.45 (8.67)123.72 (9.03)0.000
      Significant at P value ≤ 0.05.
      1.82 (1.06 to 2.59)0.002
      Significant at P value ≤ 0.05.
      3.09 (1.84 to 4.34)0.000
      Significant at P value ≤ 0.05.
      1.27 (0.17 to 2.37)0.046
      Significant at P value ≤ 0.05.
      Mandibular Height (Co-Go) (mm)66.79 (7.13)68.58 (7.73)69.91 (8.3)0.000
      Significant at P value ≤ 0.05.
      1.8 (0.96 to 2.63)0.021
      Significant at P value ≤ 0.05.
      3.12 (1.49 to 4.75)0.000
      Significant at P value ≤ 0.05.
      1.32 (−0.08 to 2.73)0.116
      Mandibular Body Length (Go-Gn) (mm)66.45 (4.55)66.79 (4.64)68 (4.51)0.000
      Significant at P value ≤ 0.05.
      0.33 (−0.34 to 1.01)0.4871.55 (1.07 to 2.02)0.000
      Significant at P value ≤ 0.05.
      1.21 (0.61 to 1.82)0.000
      Significant at P value ≤ 0.05.
      Effective Midface Length (Co-A) (mm)86.61 (5.16)87.55 (5.86)88.14 (6.78)0.001
      Significant at P value ≤ 0.05.
      0.95 (0.28 to 1.61)0.040
      Significant at P value ≤ 0.05.
      1.54 (0.72 to 2.35)0.000
      Significant at P value ≤ 0.05.
      0.59 (−0.23 to 1.41)0.275
      Maxillomandibular Differential (Co-Gn - Co-A) (mm)34.02 (4.78)34.9 (5.24)35.58 (5.39)0.004
      Significant at P value ≤ 0.05.
      0.88 (0.38 to 1.38)0.1321.56 (0.48 to 2.64)0.003
      Significant at P value ≤ 0.05.
      0.68 (−0.37 to 1.72)0.297
      A, A point; CI, confidence interval; Co, condyle; Gn, gnathion; Go, gonion; SD, standard deviation.
      a Significant at P value ≤ 0.05.

      3.4 Comparison between controls, P5, and P10

      There were no significant differences (P > 0.05) in the parameters between controls, P5, and P10 (Table 4). The percentage of patients with C2, C3, and C4 concavity classified as deep in controls, P5, and P10 was not different (Supplemental Table 4a). The agreement between controls and P5 or P10 was high for C2 and C3 concavity and was moderate for C4 concavity (Supplemental Table 4b).
      Table 4Difference between lateral cephalogram generated at 0 rotation (Controls, n = 40) and Pitch rotation of 5° (P5, n = 40) and pitch rotation of 10° (P10, n = 40)
      ParametersControlsP5P10Overall P valueP5- controlsP10- controlsP10- P5
      Mean (SD)Mean (SD)Mean (SD)P value
      Significant at P value ≤ 0.05.
      Mean (95% CI)P value
      Significant at P value ≤ 0.05.
      Mean (95% CI)P value
      Significant at P value ≤ 0.05.
      Mean (95% CI)P value
      Significant at P value ≤ 0.05.
      C2 Concavity (mm)1.42 (0.73)1.42 (0.7)1.38 (0.72)0.7620 (−0.11 to 0.11)0.999−0.03 (−0.15 to 0.08)0.787−0.03 (−0.12 to 0.06)0.813
      C3 Concavity (mm)1.21 (0.78)1.22 (0.73)1.2 (0.74)0.9580.01 (−0.11 to 0.13)0.985−0.01 (−0.12 to 0.1)0.991−0.02 (−0.15 to 0.11)0.954
      C4 Concavity (mm)0.96 (0.65)1 (0.66)1.01 (0.78)0.7240.04 (−0.08 to 0.15)0.8060.04 (−0.08 to 0.17)0.7340.01 (−0.11 to 0.13)0.991
      C3 Base-Anterior Ratio (%)106.72 (27.39)108.09 (26.31)107.82 (24.4)0.6891.37 (−2.11 to 4.85)0.6941.1 (−2.79 to 5)0.789−0.27 (−2.98 to 2.44)0.986
      C3 Posterior-Anterior Ratio (%)110.79 (18.12)110.26 (15.96)111.58 (13.57)0.725−0.53 (−4.21 to 3.15)0.9450.79 (−3.26 to 4.85)0.8821.32 (−0.61 to 3.25)0.706
      C4 Base-Anterior Ratio (%)113.55 (22.27)116.79 (23.67)112.96 (23.5)0.0693.24 (−0.75 to 7.24)0.160−0.58 (−4.44 to 3.27)0.941−3.83 (−6.45 to −1.2)0.081
      C4 Posterior-Anterior Ratio (%)116.54 (16.52)119.37 (16.57)116.75 (15.57)0.2502.83 (−0.88 to 6.53)0.2930.21 (−3.66 to 4.08)0.993−2.61 (−6.42 to 1.19)0.349
      Effective Mandibular Length (Co-Gn) (mm)120.63 (7.86)120.7 (8.29)120.79 (8.17)0.8010.07 (−0.41 to 0.55)0.9560.16 (−0.38 to 0.7)0.7850.09 (−0.32 to 0.5)0.924
      Effective Mandibular Height (Co-Go) (mm)66.79 (7.13)66.63 (7.04)67.08 (6.67)0.132−0.16 (−0.5 to 0.18)0.7600.29 (−0.25 to 0.84)0.3980.45 (−0.01 to 0.92)0.117
      Mandibular Body Length (Go-Gn) (mm)66.45 (4.55)66.24 (4.52)66.29 (4.31)0.526−0.22 (−0.66 to 0.23)0.522−0.16 (−0.5 to 0.17)0.6890.05 (−0.36 to 0.46)0.962
      Midface Length (Co-A) (mm)86.61 (5.16)86.84 (5.44)86.94 (5.66)0.4280.23 (−0.25 to 0.71)0.6480.33 (−0.26 to 0.92)0.4120.1 (−0.38 to 0.58)0.921
      Maxillomandibular Differential (Co-Gn - Co-A) (mm)34.02 (4.78)33.86 (5.05)33.85 (4.95)0.724−0.16 (−0.61 to 0.29)0.778−0.17 (−0.67 to 0.33)0.754−0.01 (−0.52 to 0.5)0.999
      A, A point; CI, confidence interval; Co, condyle; Gn, gnathion; Go, gonion; SD, standard deviation.
      a Significant at P value ≤ 0.05.

      3.5 Comparison between controls, YRP5, and YRP10

      Both YPR5 and YRP10 showed significantly increased (P < 0.05) C2, C3, and C4 concavity, ML, MH, MBL, and MMD (P < 0.05) compared with controls. YRP10 showed significantly increased (P < 0.05) C2 concavity and MMD; and significantly decreased (P < 0.05) C3 PAR, C4 PAR, and MFL than YRP5 (P < 0.05, Table 5).
      Table 5Difference between lateral cephalogram generated at 0 rotation (Controls, n = 40) and yaw, roll, and pitch rotation of 5° (Y5R5P5, n = 40) and yaw, roll, and pitch rotation of 10° (Y10R10P10, n = 40)
      ParametersControlsY5R5P5Y10R10 P10P value overall
      Significant at P value ≤ 0.05.
      Y5R5 P5- controlsY10R10 P10- ControlsY10R10P10- Y5R5P5
      Mean (SD)Mean (SD)Mean (SD)Mean (95% CI)P value
      Significant at P value ≤ 0.05.
      Mean (95% CI)P value
      Significant at P value ≤ 0.05.
      Mean (95% CI)P value
      Significant at P value ≤ 0.05.
      C2 Concavity (mm)1.42 (0.73)1.66 (0.82)1.84 (0.73)0.000
      Significant at P value ≤ 0.05.
      0.24 (0.13 to 0.34)0.001
      Significant at P value ≤ 0.05.
      0.42 (0.29 to 0.56)0.000
      Significant at P value ≤ 0.05.
      0.19 (0.05 to 0.33)0.009
      Significant at P value ≤ 0.05.
      C3 Concavity (mm)1.21 (0.78)1.52 (0.74)1.6 (0.9)0.000
      Significant at P value ≤ 0.05.
      0.31 (0.21 to 0.41)0.000
      Significant at P value ≤ 0.05.
      0.4 (0.22 to 0.57)0.000
      Significant at P value ≤ 0.05.
      0.08 (−0.08 to 0.25)0.486
      C4 Concavity (mm)0.96 (0.65)1.34 (0.65)1.33 (0.73)0.000
      Significant at P value ≤ 0.05.
      0.38 (0.25 to 0.51)0.000
      Significant at P value ≤ 0.05.
      0.37 (0.24 to 0.5)0.000
      Significant at P value ≤ 0.05.
      −0.01 (−0.15 to 0.13)0.993
      C3 Base-Anterior Ratio (%)106.72 (27.39)102.51 (24.16)99.47 (22.28)0.007
      Significant at P value ≤ 0.05.
      −4.21 (−7.73 to −0.69)0.151−7.25 (−12.07 to −2.43)0.005
      Significant at P value ≤ 0.05.
      −3.04 (−8.12 to 2.04)0.367
      C3 Posterior-Anterior Ratio (%)110.79 (18.12)109.73 (18.98)102.25 (18.36)0.004
      Significant at P value ≤ 0.05.
      −1.06 (−5.28 to 3.17)0.921−8.54 (−14.84 to −2.24)0.007
      Significant at P value ≤ 0.05.
      −7.48 (−13.27 to −1.7)0.020
      Significant at P value ≤ 0.05.
      C4 Base-Anterior Ratio (%)113.55 (22.27)109.82 (18.31)105.62 (20.23)0.015
      Significant at P value ≤ 0.05.
      −3.73 (−8.58 to 1.13)0.343−7.92 (−14.31 to −1.54)0.010
      Significant at P value ≤ 0.05.
      −4.2 (−8.89 to 0.5)0.260
      C4 Posterior-Anterior Ratio (%)116.54 (16.52)118.4 (16.69)106.22 (12.5)0.000
      Significant at P value ≤ 0.05.
      1.86 (−4.8 to 8.52)0.809−10.31 (−16.7 to −3.93)0.003
      Significant at P value ≤ 0.05.
      −12.18 (−17.18 to −7.17)0.000
      Significant at P value ≤ 0.05.
      Effective Mandibular Length (Co-Gn) (mm)120.63 (7.86)122.61 (7.6)122.06 (8.74)0.000
      Significant at P value ≤ 0.05.
      1.99 (1.13 to 2.84)0.000
      Significant at P value ≤ 0.05.
      1.43 (0.36 to 2.5)0.007
      Significant at P value ≤ 0.05.
      −0.55 (−1.41 to 0.3)0.454
      Mandibular Height (Co-Go) (mm)66.79 (7.13)68.42 (7.05)68.85 (7.98)0.000
      Significant at P value ≤ 0.05.
      1.63 (0.74 to 2.52)0.003
      Significant at P value ≤ 0.05.
      2.06 (1.04 to 3.08)0.000
      Significant at P value ≤ 0.05.
      0.43 (−0.55 to 1.41)0.639
      Mandibular Body Length (Go-Gn) (mm)66.45 (4.55)67.92 (4.05)67.79 (4.21)0.000
      Significant at P value ≤ 0.05.
      1.46 (0.73 to 2.19)0.000
      Significant at P value ≤ 0.05.
      1.33 (0.53 to 2.14)0.001
      Significant at P value ≤ 0.05.
      −0.13 (−0.74 to 0.48)0.932
      Effective Midface Length (Co-A) (mm)86.61 (5.16)87.64 (4.92)85.11 (6.14)0.000
      Significant at P value ≤ 0.05.
      1.03 (0.19 to 1.86)0.059−1.5 (−2.5 to −0.5)0.003
      Significant at P value ≤ 0.05.
      −2.52 (−3.37 to −1.68)0.000
      Significant at P value ≤ 0.05.
      Maxillomandibular Differential (Co-Gn - Co-A) (mm)34.02 (4.78)34.98 (5.02)36.95 (5.24)0.000
      Significant at P value ≤ 0.05.
      0.96 (0.25 to 1.67)0.009
      Significant at P value ≤ 0.05.
      2.93 (2.31 to 3.55)0.000
      Significant at P value ≤ 0.05.
      1.97 (1.39 to 2.55)0.000
      Significant at P value ≤ 0.05.
      A, A point; CI, confidence interval; Co, condyle; Gn, gnathion; Go, gonion; SD, standard deviation.
      a Significant at P value ≤ 0.05.
      The percentage of patients with C2, C3, and C4 concavity classified as deep was higher in YRP10 than YRP5 and controls (Supplemental Table 5a). The Fleiss’ kappa coefficient showed moderate agreement between controls, YRP5 for C2, and C3 concavity (Supplemental Table 5b). It should be noted that there was high individual variation for the measured parameters between the different rotations and controls.
      The intrarater reliability and interrater reliability showed good agreement with intraclass correlation coefficients >0.8 for all measurements (Supplemental Table 6).

      4. Discussion

      Our null hypothesis was rejected, as we observed that rotational errors in yaw and roll, but not pitch, led to differences in the CVM staging. Specifically, this study showed that rotational errors can lead to an exaggeration of the CVM indicators; therefore, rotational errors of head position have a significant effect on the diagnostic accuracy of the cephalogram. Some studies have suggested having a standardized position of patients, whereas others have suggested using natural head position [
      • Cooke M.S.
      • Wei S.H.
      The reproducibility of natural head posture: a methodological study.
      ,
      • Leitão P.
      • Nanda R.S.
      Relationship of natural head position to craniofacial morphology.
      ]. Clinically, the head orientation for recording the lateral cephalogram is often based on the midsagittal plane and the obitale, or on the left and right porion. Therefore, the orbitale line from the frontal view and Frankfort Horizontal plane (orbitale to porion) were used to orient the CBCT and investigate Cervical Vertebrae Maturation Indicator (CVMI) errors in lateral cephalograms due to errors in the head orientation induced in this way. In addition, in certain clinical cases, the midsagittal plane is difficult to set due to facial asymmetry. This study shows how much head orientation errors of different degrees cause CVMI errors, which can be useful in the diagnosis of CVMI in such clinical situations.
      A CBCT shows the 3D structures of the patient and is a 1:1 representation of the patient's hard tissue [
      • Marmulla R.
      • Wörtche R.
      • Mühling J.
      • Hassfeld S.
      Geometric accuracy of the NewTom 9000 Cone Beam CT.
      ]. Thus, to analyze the accuracy of the lateral cephalogram in evaluating CVM, the CBCT and controls were compared. Table 2 shows that the controls had significantly higher concavity on C2 and C4 compared with CBCT (Table 1). In addition, the percentage of patients that were identified as having deep concavity in C3 and C4 were almost double in controls compared with CBCT (Supplemental Table 1c). In addition, the maxillomandibular linear parameters were significantly greater in controls. This may result because of the errors of magnification, distortion, and difficulty in landmark identification due to the superimposition of structures in the lateral cephalogram [
      • Baccetti T.
      • Franchi L.
      • McNamara Jr., J.A.
      An improved version of the cervical vertebral maturation (CVM) method for the assessment of mandibular growth.
      ,
      • Chen Y.J.
      • Chen S.K.
      • Yao J.C.
      • Chang H.F.
      The effects of differences in landmark identification on the cephalometric measurements in traditional versus digitized cephalometry.
      ]. As the CVM stage was developed and calibrated using lateral cephalograms, it should be realized that the stage of CVM when assessed on CBCT appears lower than the actual CVM stage that is visible on lateral cephalogram. Clinicians should adjust their evaluation when viewing CBCT images and should consider the patient is at a later stage than what may appear in the CBCT.
      A comparison of controls with Y5 and Y10 showed that at Y5, no difference in cervical vertebral concavities was noted (Table 2). This shows that minor rotations in yaw do not adversely affect the assessment of CVM; however, when evaluating Y5, a significant decrease was found in MFL. To evaluate whether increasing the amount of yaw rotation has any effect on CVM assessment, Y10 cephalogram was compared with controls. C4 concavity was significantly higher in Y10 (Table 2) and the percentage of patients classified as having deep C2, C3, and C4 concavity was higher in Y10 (Supplemental Table 2a). This shows that the assessment of the CVM stage may be overestimated if the head position is rotated by 10° in the horizontal plane. In addition, Y10 also showed significantly decreased ML, MH, MBL, and MFL. This may be due to difficulty in accurately identifying the condylion and A point with an increased rotation of the head in the horizontal plane [
      • Nikneshan S.
      • Mohseni S.
      • Nouri M.
      • Hadian H.
      • Kharazifard M.J.
      The effect of emboss enhancement on reliability of landmark identification in digital lateral cephalometric images.
      ].
      The assessment of the effects of the head rotation in the frontal plane showed that C2, C3, and C4 concavity were not affected by 5° of roll rotation but were affected by 10°. It has been reported that the rotation of the head may lead to changes in posterior-anterior cephalometric measurements [
      • Major P.W.
      • Johnson D.E.
      • Hesse K.L.
      • Glover K.E.
      Effect of head orientation on posterior anterior cephalometric landmark identification.
      ]. In this study, we found that head rotation also affects the lateral cephalometric measurements of CVM and maxillomandibular linear parameters. There was an increasing trend in the depth of cervical vertebral concavities measurement from controls to R5 to R10. Thus, proper head positioning becomes critical when recording the lateral cephalogram, as a greater deviation from the normal leads to decreased diagnostic accuracy. The results showed that ML, MH, and MFL were significantly higher in the R5 group (Table 4) and even higher in R10. These findings reflect that the errors in positioning in the frontal plane (roll) lead to different types of errors on lateral cephalometric measurements than the horizontal plane (yaw). In roll rotation, the mandibular and maxillary parameters were overestimated, whereas in cephalograms with yaw rotation, these parameters were underestimated. In contrast to Y5, Y10, R5, and R10, there was no significant difference in the CVM indicators between P5, P10, and control measurements (Table 4). Hence, it can be concluded that the rotation in the sagittal plane does not affect the CVM accuracy to a significant degree.
      To compare the effects of combined rotations of the head in different planes, YRP5 and YRP10 were compared with controls. A significant difference in C2, C3, and C4 concavity in YRP5 was observed when compared with controls. In addition, the ML, MH, MBL, and MMD were significantly higher in YRP5 (Table 5). These findings show that more significant changes occur with 5° of multiplane rotation than with single plane rotation. When YRP10 was evaluated, all the parameters were found to be significantly different from controls. The percentage of patients identified as having deep concavity in C2, C3, and C4 was higher in YRP10 than YRP5, which was in turn higher than controls (Supplemental Table 5a). Thus, it can be concluded that multiplane rotations caused higher inaccuracies than single plane rotations. In addition, a higher degree of rotation (10°) led to greater inaccuracies than mild rotations (5°), leading to overestimation of the CVM stage. The 95% confidence interval indicates that the individual variations were high when comparing the parameters between controls and lateral cephalograms with different rotations (Table 1, Table 2, Table 3, Table 4, Table 5). Even though the mean differences were statistically significant, it should be noted that the effects are variable considering the large confidence intervals.
      There were some limitations in this study, such as the retrospective study design and the accuracy of the measurements being limited by the voxel size of the CBCT. Another limitation was that for the purpose of comparison, the left and right values from the CBCTs were averaged; this may lead to errors in patients with skeletal asymmetry. In this study, we have measured a model of head movement that may emulate head movement during patient positioning; however, the exact relationship between the model and patient positioning has not been established. Nevertheless, there is no literature available in relation to the amount of positioning errors in pitch, roll, and yaw and its effect on CVM staging. Thus, the effect of the rotations in different planes and different magnitudes was assessed. The results of this study indicate that the assessment of CVM from the lateral cephalograms with positioning errors might not be accurate. Therefore, the CVM assessment from such cephalometric radiographs should be used with caution and the tendency to overestimate the CVM stage with positioning errors in radiographs must be taken into consideration while treatment planning.

      5. Conclusions

      • Direct measurements from CBCT images are likely to lead to an underestimation of the CVM stage when compared with those estimated from artificial lateral cephalograms, derived from those CBCT images.
      • Analysis of CVM stages in lateral cephalograms with positioning errors in the horizontal plane (yaw) and frontal plane (roll) may be inaccurate. The rotational errors in the horizontal plane (yaw) will likely lead to an underestimation of maxillomandibular linear measurements, whereas the rotational errors in the vertical plane (roll) will likely lead to an overestimation of the measurements.
      • Positional errors in the sagittal plane (pitch) are unlikely to affect the accuracy of CVM measurements significantly.
      • There may be more inaccuracies in measurements relating to CVM stages and maxillomandibular parameters in lateral cephalograms generated with multiplane rotations than single plane rotations.
      • The lateral cephalograms with a higher degree of rotation (10°) may lead to more inaccuracies in CVM stages when compared with lateral cephalograms with a mild rotation of (5°).
      • There is wide individual variation in CVM staging and measurements taken from the control and rotation groups. Hence, sound clinical judgment must be used when interpreting the results of this study.

      Supplemental materials

      Supplemental material accompanying this article can be found in the online version as a hyperlink at https://doi.org/10.1016/j.ejwf.2020.09.006 or, if a hard copy of article, at http://ees.elsevier.com/jwfo/(select volume, issue, and article).

      Appendix A. Supplementary Data

      • Supplemental Fig. 2

        Identification of landmarks in the 3D slices view. (A), (B), and (C): identification of condylion in the coronal slice, sagittal slice, and axial slice, respectively.

      References

        • Lewis A.B.
        • Garn S.M.
        The relationship between tooth formation and other maturation factors.
        Angle Orthod. 1960; 30: 70-77
        • Hunter C.J.
        The correlation of facial growth with body height and skeletal maturation at adolescence.
        Angle Orthod. 1966; 36: 44-54
        • Fishman L.S.
        Maturational patterns and prediction during adolescence.
        Angle Orthod. 1987; 57: 178-193
        • Franchi L.
        • Baccetti T.
        New emphasis on the role of mandibular skeletal maturity in dentofacial orthopedics.
        in: McNamara J.A. The Enigma of the Vertical Dimension. Craniofacial Growth Series. 36. Center for Human Growth and Development, The University of Michigan, Ann Arbor, MI2000: 253-275
        • Baccetti T.
        • Franchi L.
        • Toth L.R.
        • McNamara Jr., J.A.
        Treatment timing for Twin-block therapy.
        Am J Orthod Dentofacial Orthop. 2000; 118: 159-170
        • O'Reilly M.T.
        • Yanniello G.J.
        Mandibular growth changes and maturation of cervical vertebrae--a longitudinal cephalometric study.
        Angle Orthod. 1988; 58: 179-184
        • Franchi L.
        • Baccetti T.
        • McNamara Jr., J.A.
        Mandibular growth as related to cervical vertebral maturation and body height.
        Am J Orthod Dentofacial Orthop. 2000; 118: 335-340
        • Beit P.
        • Peltomäki T.
        • Schätzle M.
        • Signorelli L.
        • Patcas R.
        Evaluating the agreement of skeletal age assessment based on hand-wrist and cervical vertebrae radiography.
        Am J Orthod Dentofacial Orthop. 2013; 144: 838-847
        • Ball G.
        • Woodside D.
        • Tompson B.
        • Hunter W.S.
        • Posluns J.
        Relationship between cervical vertebral maturation and mandibular growth.
        Am J Orthod Dentofacial Orthop. 2011; 139: e455-e461
        • Major P.W.
        • Johnson D.E.
        • Hesse K.L.
        • Glover K.E.
        Effect of head orientation on posterior anterior cephalometric landmark identification.
        Angle Orthod. 1996; 66: 51-60
        • Shim J.J.
        • Heo G.
        • Lagravère M.O.
        Assessment of skeletal maturation based on cervical vertebrae in CBCT.
        Int Orthod. 2012; 10: 351-362
        • Joshi V.
        • Yamaguchi T.
        • Matsuda Y.
        • Kaneko N.
        • Maki K.
        • Okano T.
        Skeletal maturity assessment with the use of cone-beam computerized tomography.
        Oral Surg Oral Med Oral Pathol Oral Radiol. 2012; 113: 841-849
        • Kim H.J.
        • Hong M.
        • Park H.S.
        Analysis of dental compensation in patients with facial asymmetry using cone-beam computed tomography.
        Am J Orthod Dentofacial Orthop. 2019; 156: 493-501
        • Ackerman J.L.
        • Proffit W.R.
        • Sarver D.M.
        • Ackerman M.B.
        • Kean M.R.
        Pitch, roll, and yaw: describing the spatial orientation of dentofacial traits.
        Am J Orthod Dentofacial Orthop. 2007; 131: 305-310
        • Baccetti T.
        • Franchi L.
        • McNamara Jr., J.A.
        An improved version of the cervical vertebral maturation (CVM) method for the assessment of mandibular growth.
        Angle Orthod. 2002; 72: 316-323
        • Green P.
        • Macleod C.J.
        SIMR: An R package for power analysis of generalized linear mixed models by simulation.
        Methods Ecol Evol. 2016; 7: 493-498
        • Cooke M.S.
        • Wei S.H.
        The reproducibility of natural head posture: a methodological study.
        Am J Orthod Dentofacial Orthop. 1988; 93: 280-288
        • Leitão P.
        • Nanda R.S.
        Relationship of natural head position to craniofacial morphology.
        Am J Orthod Dentofacial Orthop. 2000; 117: 406-417
        • Marmulla R.
        • Wörtche R.
        • Mühling J.
        • Hassfeld S.
        Geometric accuracy of the NewTom 9000 Cone Beam CT.
        Dentomaxillofac Radiol. 2005; 34: 28-31
        • Chen Y.J.
        • Chen S.K.
        • Yao J.C.
        • Chang H.F.
        The effects of differences in landmark identification on the cephalometric measurements in traditional versus digitized cephalometry.
        Angle Orthod. 2004; 74: 155-161
        • Nikneshan S.
        • Mohseni S.
        • Nouri M.
        • Hadian H.
        • Kharazifard M.J.
        The effect of emboss enhancement on reliability of landmark identification in digital lateral cephalometric images.
        Iran J Radiol. 2015; 12: e19302