This study aimed to evaluate the 3-dimensional (3D) mandibular dental changes over 42 years using the registration of digital models.
The sample comprised digital dental models of 8 untreated subjects (4 males and 4 females) with normal occlusion measured longitudinally at ages 17 years (T1) and 60 years (T2). Using 13 landmarks placed on the mucogingival junction, we registered the T2 model on the T1 model. Three-dimensional changes in the position of the landmarks on the buccal cusp tip of the posterior teeth and incisal edge of the central incisors were measured by 2 examiners. Registration and measurements were performed using SlicerCMF (version 3.1; www.slicer.org ) software. Intra- and interrater agreements were evaluated using intraclass correlation coefficients and the Bland-Altman method. One-sample t tests were used for evaluating interphase 3D dental changes ( P <0.05).
Adequate intra- and interrater reproducibility was found. From T1 to T2, the mandibular teeth showed significant 3D positional changes. A significant dental eruption relative to the mucogingival junction was observed for the anterior and posterior teeth. Anteroposterior movements of mandibular teeth were not significant except for the right molar that drifted mesially. Transverse movements included slight lingual tipping at canines and premolars regions.
Dental changes in untreated normal occlusion were very slight from early to mature adulthood. The eruption of the mandibular teeth was the most consistent finding. A tendency for mesial movement of molars and lingual movement of first premolars and canines was observed in the mandible during the aging process.
We evaluated mandibular dental changes using the registration of digital models.
The sample comprised 8 normal occlusion dental models taken at ages 17 years and 60 years.
After 42 years, mandibular teeth showed small but significant 3-dimensional positional changes.
Skeletal and dentoalveolar changes during the growth of the face are continuous processes occurring throughout life. , Maturational changes take place both in the maxillary and mandibular dental arches. , , Mandibular arch changes are of particular interest to the orthodontist because of the frequent occurrence of late incisor crowding, , a common complaint of patients long after orthodontic treatment has been completed.
There have been several previous studies that have evaluated long-term changes in the dentition of untreated subjects. Sinclair and Little found a significant decrease in mandibular arch length, intercanine width, and intermolar width in a sample of 65 untreated subjects evaluated in the mixed, early permanent dentition, and early adulthood. A significant increase in the mandibular incisor irregularity was observed in the permanent dentition period.
Bishara et al evaluated dental models derived from a sample with normal occlusion at ages 25 years and 46 years. A significant decrease in mandibular intercanine width and arch length was found. In addition, significant mandibular dental crowding took place during the 20 years of follow-up. Carter and McNamara Jr examined an untreated sample of subjects aged from 16 years to 48 years. A significant decrease in arch width, arch depth, and arch perimeter was found for the mandibular arch. Mandibular incisor irregularities showed a significant increase of 1.4 mm over a 30-year follow-up.
Thilander analyzed a sample of subjects with normal occlusion, aged from 5 years to 31 years. A decrease of 4 mm in the arch perimeter was found in the mandible. After the age of 16 years, a continuous decrease in intercanine width was found in the mandibular arch, resulting in anterior crowding, especially in the mandible. Tsiopas et al evaluated dental models of untreated Swedish dentists from the ages of 20 years to 60 years. The authors found a significant increase in the irregularity index of Little in the mandibular arch and a decrease of arch length and intercanine distance in both dental arches.
Miranda et al examined dental models of subjects between the ages of 13 years and 62 years with normal occlusion. The authors found significant mandibular incisor crowding occurring only in subjects without permanent tooth loss. Massaro et al evaluated the maturational changes in normal occlusion subjects throughout 40 years. Intercanine width and arch perimeter decreased from 18 to 60 years. Incisor crowding was found only in the mandible. The authors suggested that future studies with dental model superimpositions were necessary to elucidate if the decrease in intercanine width occurs because of mesial or lingual canine displacements.
The superimposition of digital dental models enables quantification and visual analysis of 3-dimensional (3D) tooth movements and arch changes over time. To achieve this goal, accurate registration of digital models from different time points using reliable reference areas is required. Previous studies on the superimposition of digital dental models for tooth movement analysis were limited to the maxillary arch, primarily using the palatal rugae as reference. Considering the technological development, a search for stable areas or points of reference for the mandibular dental arch was needed.
Park et al proposed a method of mandibular dental model superimposition based on cone-beam computed tomography (CBCT) surface registration of the mandible in nongrowing patients. First, CBCT images of digital models were superimposed at corresponding time points using the best-fit method. Subsequently, the CBCT images were superimposed at the mandibular basal bone and posterior ramus regions. After deleting the tooth images from the CBCT, landmarks were assigned to digital dental models allowing reproducible measurements of 3D dental changes.
An et al evaluated 4 regions of interest on the mandibular alveolar ridge and reported that the mandibular tori were stable areas for the superimposition of the digital dental model. The authors found that the horizontal and vertical movements of the central incisors and first molars measured on the superimposed models were similar to 2-dimensional cephalometric changes.
More recently, Schmidt et al developed a method for dental model superimposition combining local rigid tooth surface registration with a nonimage based method driven by biomechanical models. The authors reported that their methods were applicable to both maxilla and mandible for orthodontic tooth movement analysis and that growth-related changes are ignored. Ioshida et al validated the mucogingival junction as a reference for mandibular digital dental model registration in a short-term assessment of 48 weeks. Voxel-based registration of mandibular CBCT images was compared with registered dental models on the mucogingival junction, and no significant differences were found.
Although several studies have investigated the maturational process in the mandibular arch using dental casts, the 3D direction of mandibular teeth movements that occur with aging, and affect the stability of orthodontic correction, require further investigation. The superimposition of digital models could potentially elucidate the individual teeth movement over time. Therefore, the purpose of this study was to evaluate the 3D mandibular dental changes during aging using the registration of digital models on the mucogingival junction. Maturational mandibular tooth movement from ages 17 years to 60 years was measured 3-dimensionally in normal occlusion subjects.
Material and methods
This study was approved by the institutional research ethical committee of Bauru Dental School, University of São Paulo (Process no. 43931915.4.0000.5417).
Written patient consent forms were obtained. Sample size calculation considered a test power of 80%, an alpha error of 5%, a minimum difference of 1 mm to be detected, and a standard deviation of 0.8 mm derived from preliminary statistics. A sample of 7 patients was necessary.
The sample of this study consisted of dental models of untreated subjects with normal occlusion. The original sample was composed of 82 white subjects, recruited in the 1960s and 1970s. Dental models obtained at age 17 years (T1) were used. All subjects were recalled in 2015 and 2016 (T2). From the initial sample, 38 subjects could be contacted, 36 could not be found, and 8 had died. Eleven subjects did not agree to return at T2. Twenty-seven subjects initially were enrolled in the sample.
The exclusion criteria included a history of previous orthodontic treatment, tooth loss of any tooth mesial to second molars from T1 to T2, absence of prosthetic rehabilitation, and inadequate dental models. From the subjects enrolled, 3 had inadequate T1 study models, and 16 subjects were excluded because of tooth loss and prosthetic rehabilitation at T2. The final sample consisted of 8 subjects with normal occlusion (4 males and 4 females).
The dental models taken at ages 17 years and 60 years were scanned using an R700 3D Scanner (3Shape, Copenhagen, Denmark). The maxillary and mandibular dental models were scanned and stored as Stereolithography files (.STL) were converted to visualization toolkit (.VTK) mesh files, using SlicerCMF (version 3.1; www.slicer.org ). Registration and measurements consisted of the following steps:
Model orientation: The maxillary and mandibular T1 dental models were oriented in centric occlusion using the 3D coordinate system in the transforms tool of SlicerCMF software. Using the maxillary occlusal perspective, we positioned the midpalatal raphe coincident with the anteroposterior yellow line (sagittal plane). In the same view, the second palatal rugae was moved to coincide with the superior-inferior green line (coronal plane). In the right-side view, the occlusal plane, defined by a line passing through the maxillary first molar mesiobuccal cusp tip and maxillary canine cusp tip, was placed on the right-left red line (axial plane). On the model’s frontal view, the cusp tips of the right and left canines were positioned on the red line (axial plane). After orientation, the maxillary dental model was eliminated, and all the following steps were performed only on the mandibular dental models.
Model approximation: Model approximation was conducted in 2 steps. First, the T2 mandibular dental model was approximated to the T1 model by placing 6 corresponding landmarks in the T2 and T1 3D digital dental models using the CMF registration module in SlicerCMF. The landmarks were placed on the tip of the mesiobuccal cusp of the first molar, the buccal cusp of the first premolar and the canine. All landmarks were placed bilaterally. The Q3DC module of the SlicerCMF software displayed the x-, y-, and z-coordinates for each landmark. Using the x-, y-, and z-coordinates of T1 as a reference, the software changed the spatial position of T2 to match the T1 coordinates. As a result, the T2 and T1 mandibular models were approximated by the superimposition of the corresponding landmarks facilitating the next step. The second step consisted of manual approximation of the mucogingival junction of the T2 mandibular model to the mucogingival junction of the T1 model.
Registration: On both T1 and T2 mandibular models, 13 landmarks were placed on the mucogingival junction between each permanent tooth from the distal aspect of the mandibular right first molars to the distal aspect of the left homologous tooth ( Fig 1 ). Using the CMF registration module in SlicerCMF, the T2 mandibular model was registered relative to the T1 model by matching the coordinates of the corresponding landmarks. The software automatically calculated the best fit for the simultaneous superimposition of the corresponding landmarks.
Three-dimensional measurements: Using the Q3DC module of the SlicerCMF software, we placed the landmarks on T1 and on the registered T2 models at the tip of the mesiobuccal cusp of the first molars, buccal cusp of the first premolars and canines, bilaterally, and on the mesial angle of the incisal edge of the left central incisor ( Fig 2 ). Visualization of the occlusal view of the 2 time points of each pair of mandibular models was available, side-by-side on the screen while placing the dental landmarks. Differences between T1 to T2 were measured considering the complete 3D displacement and the changes in x-, y-, and z-coordinates. Anterior, superior, and lateral displacements had positive values. Posterior, inferior, and medial displacements had negative values.
Statistical analysis was carried out using SPSS (version 21.0; SPSS, Chicago, Ill). Steps 1 to 4 were performed by 2 examiners (D.G and F.M). The first examiner repeated the steps twice with a 15-day interval. Descriptive statistics included the mean and standard deviation of 3D tooth displacements from T1 to T2, comprising a 42-year interval. Intra- and interrater agreements were calculated using a 1-way random intraclass correlation coefficient and Bland-Altman plot. Intraclass correlation coefficient values from 0.75 to 1, from 0.6 to 0.74, from 0.4 to 0.59, and less than 0.4 were considered excellent, good, fair, and poor agreements, respectively. A 1-sample t test was used for evaluating interphase 3D dental changes ( P <0.05).
|Tooth||Measurement||3D changes (mm)||Intrarater agreement ICC||Interrater agreement ICC||Intrarater differences (Bland-Altman)||Interrater differences (Bland-Altman)|
|Right first molar||R-L||−0.11||0.80||0.72||0.95||0.90||0.05||0.33||−0.60||0.70||−0.04||0.56||−1.14||1.06|
|Right first premolar||R-L||−0.58||0.54||0.01 ∗||0.93||0.74||0.12||0.20||−0.27||0.51||0.01||0.59||−1.15||1.16|
|Right canine||R-L||−0.73||0.69||0.01 ∗||0.95||0.90||−0.12||0.18||−0.46||0.23||−0.02||0.38||−0.77||0.73|
|Central incisor||R-L||−0.20||0.22||0.03 ∗||0.84||0.32||0.00||0.18||−0.36||0.35||0.12||0.38||−0.63||0.87|
|Left first premolar||R-L||−0.39||0.63||0.11||0.92||0.58||−0.05||0.22||−0.48||0.39||0.29||0.37||−0.44||1.03|
|Left first molar||R-L||−0.24||0.87||0.45||0.91||0.90||0.15||0.30||−0.43||0.73||−0.04||0.39||−0.80||0.72|