Reprodutibilidade, confiabilidade e validade de um modelo biomecânico tridimensional de análise de vantagem mecânica dos músculos envolvidos no movimento mandibular

Abstract

The objective of this study was to evaluate the reproducibility and the errors of a threedimensional mechanical advantage model of mandibular closing muscles in Cone beam computed tomography (CBCT) images, as well as to verify the reliability and validity of this analysis in relation to the conventional two-dimensional model. A total of 30 CBCT images were randomly selected. For the construction of the threedimensional model, the Sagittal, Mandibular and Frankfort reference planes were identified, as well as the landmarks landmarks Foramen spinosum (Fs), Nasion (N), Gnation (Gn), Orbital (Or), Porion (Po), Condylion (Co), Coronoid (Cor), Gonion (Go), Isthmus (Ne), Pterygoid (Pt), Ramus (Ra), Zygomatic tuberosity, Sutures temporomalar and maxilomalar, Incisal (In), and Molar (Mo). For the two-dimensional model, the tomographic sections were grouped into anteroposterior constructions. The reference planes used were Frankfort and Mandibular, and the landmarks identified were (S), Nasion (N), Gnation (Gn), Orbital (Or), Porion (Po), Condylion (Co), Coronoid (Cor), Gonion (Go), Pterygoid (Pt), Incisal (In), and Molar (Mo). From these points, drew up the lever arms for the superficial masseter, anterior and posterior deep masseter, anterior and posterior temporal, medial and lateral pterygoid muscles, as well as, their corresponding resistance arms on the incisor and molar level. The mechanical advantage for each muscle was calculated. This procedure was repeated 3 times by the same examiner and repeated once by 2 examiners. For each method, reproducibility was determined using the intra-class correlation coefficient (ICC), and errors through the average mean of absolute differences, the estimator "Method of Moments" and Dahlberg equation. Intra-examiner and inter-examiner reliability was determined by comparison of the average mean of absolute differences for corresponding variables from both models using paired Student's t test. Validity was also calculated using paired Student's t test to compare the mean measurements of the five repetitions from both models. The coordinate values corresponding to the tridimensional cephalometric points demonstrated excellent intra-examiner and interexaminer reproducibility (ICC: 0.998 – 1.000; p<0.0001). The tridimensional intraexaminer data showed errors smaller than 1.5 mm, which is considered clinically acceptable. The clinically unsatisfactory inter-examiner errors were from 1.51 mm to 5.83 mm. All variables presented excellent reproducibility, showing an ICC of 0.919 to 1,000 in the intra-examiner comparisons (p <0.0001). In the inter-examiner comparisons, the results can also be considered excellent, although with lower values (0.750 - 1,000; p<0.0001). However, the lever arm and the mechanical advantages at the molar and incisal level on the right and left sides of the lateral pterygoid muscle had ICCs below 0.500 (p<0.05). In the intra-examiner and inter-examiner comparisons, the mechanical advantage only varied from 0.00 mm to 0.01 mm and 0.01 mm to 0.05 mm, respectively. The coordinate values corresponding to the bidimensional cephalometric points demonstrated excellent intra-examiner and inter-examiner reproducibility (ICC: 0.988 – 1.000; p<0.0001). The intra-examiner tridimensional data 20 showed errors smaller than 1.5 mm, which is considered clinically acceptable. Clinically unsatisfactory inter-examiner errors were from 1.93 mm to 3.80 mm. All biomechanical variables presented excellent reproducibility, showing an ICC of 0.979 to 1,000 in intra-examiner comparisons (p<0.0001). In the inter-examiner comparisons, the results were satisfactory (ICC: 0.504-0.866; p<0.0001). Intra-examiner errors were clinically acceptable (<1.5 mm), with all mechanical advantages exhibiting an error of 0.00 to 0.01 mm. In the inter-examiner analysis, the lever and resistance arms showed errors that exceeded the limit of 1.5 mm (2.02 mm to 5.44 mm). All mechanical advantages showed an error of 0.01 to 0.08. In the assessment of reliability, intraexaminer and inter-examiner errors presented different values in general (p<0.05). However, in the intra-examiner comparison, the landmarks Nasion, Lower incisal, Porium and Molar showed a similar error in both methods. This tendency was also observed for the Gnathion, Lower incisal, Orbitale, Porium, Gonion and Molar landmarks in the inter-examiner comparison. With the exception of the lever arm and the mechanical advantages of the anterior temporal (p<0.05), almost all of the values of the biomechanical variables showed intra-examiner equivalence. The opposite occurred for all inter-examiner comparisons (p<0.05). In the comparison of the means of the measurements of the 5 repetitions for the validity determination, the points Nasion, Orbitale, Porium, Gonion, Coronoid and Molar presented at least equivalence in one of its coordinates between the three-dimensional and two-dimensional models. Although with clinically acceptable values (difference of means <0.15), almost all of the biomechanical variables showed discrepant values (p <0.05), with the exception of the Molar resistance arm (p = 0.093). Tridimensional mechanical advantage analysis in CBCT images showed high reproducibility and acceptable errors. However, its results differ in general from the two-dimensional model. This analysis can be suggested as biomechanical research and teaching tools of the masticatory system.