The present study reports the preliminary results of the randomized clinical trial registered at www.register.clinicaltrials.gov with registration number XXXXXXXXX.
The study has been conducted with two parallel groups with an allocation ratio of 1:1.
Inclusion criteria for the patients were the following:
- presence of one or two impacted maxillary canine requiring surgical exposure and orthodontic treatment.
- permanent teeth extraction-based treatment
- current or previous orthodontic treatment in the last 12 months
- current systemic disease
- current antibiotic or anti-inflammatory therapy that could possibly compromise the results.
Two kind of interventions were planned: the first group of patients received a TPA as anchorage unit for canine traction; in the second group intervention was held with a TAD as anchorage unit.
In the TAD group, an 8 mm long miniscrew (Orthoeasy, Forestadent, Pforzheim, Germany) was used as anchorage. In both groups the approach to solve the impaction was “canine first”: no anchorage preparation was performed, besides the TPA or the miniscrew. The biomechanics included a Beta-titianium cantilever applying a force of 50-60 g measured with a pen gauge. The miniscrews insertion sites varied depending on the impacted canine position, as well as the cantilever in the TPA group. The day of the surgical exposure of the canine was coincident with the beginning of traction.
All patients agreed in having 2 CBCTs in two different time points of treatment as it was described in the experimental protocol approved by the ethical committee and in the patient’s consensus form.
The first scan was taken before the surgical exposure of the canine and beginning of traction (T0), and the second one about 3 months after (108 days in the test group, 105 in the control group) (T1).
Due to variations in the CBCT image acquisition protocol in this study scans, the “Downsize” tool in Slicer was utilized to standardize the image resolution and avoid any heterogeneity of the imaging data. All scans were reformatted to a 0.5 mm3 voxel the original scans of 0.4 mm3 voxel size using SlicerCMF version 4.0 (https://sites.google.com/a/umich.edu/dentistry-image-computing/) to standardize the scan resolution and decrease the computational power and time for image registration.
Creation of the virtual model
The first step in image processing was to export the scans in Digital Imaging and Communications in Medicine (DICOM) format and then to convert them into GIPL format for image de-identification.
From the cross-sections of the volumetric data set, virtual three-dimensional models from T0 and T1 scans were created, using ITK-SNAP open-source software. This process, called segmentation, required outlining the shape of the dental arches visible in the slices, setting up a threshold of the tissues density in order to select the anatomical structure of interest.
Hence the 3D model of the canine could be isolated from the rest of the dental arch (Fig 1).
The first step in the registration process was to determine which structure would be used as a stable reference following the maxillary regional registration methods validated by Ruellas et al. . As the timespan was very short for a significant skeletal growth, and the treatment had only dental effects, the maxillary bone was considered a good structure for reference.
The registration procedure does not depend on the precision of the 3D surface models, but actually compares voxel by voxel of gray level CBCTs images, and calculates the rotation and translation parameters between the 2 time point images.
This is a fully automated process, that is simplified by a primary manual overlap of the two CBCTs using CMF registration module in SlicerCMF (https://sites.google.com/a/umich.edu/dentistry-image-computing/).
Once both images from different time points are registered, they share the same coordinate system.
Overlay of the 3D models and quantitative measuraments
The next step included the use of VAM software (VAM v. 3.7.6, Canfield 113 Scientific Inc., Fairfield, NJ) for overlaying the registered 3D models, that allowed to evaluate the displacement of the canine and to measure the distance between the tip of the cusp of the canine resulting from the CBCT at T0 and that at T1, as well as for the apex. The software allows the selection of two points and calculates distance in mm between two points (FIG 2a-Fig 2b).
Gerig et al. proposed the use of color maps generated from closest-point distances between the surfaces . The CMF tool calculates thousands of color-coded surface distances in millimeters between 3D models surface triangles at two different time points. The color maps indicate inward (blue) or outward (red) displacement between overlaid structures. An absence of changes is indicated by the green color. In this study, color-coded maps were utilized just for visualization and qualitative assesments, but not to measure canine movement (Fig 3).
The primary outcome measurement was the canine speed of movement evaluated using a voxel based superimposition of two consecutive CBCTs. The CBCTS were acquired at baseline and 3 months after starting treatment for both groups. Once the linear displacement of the canine was measured, this was divided by the observation period in weeks to obtain the ratio of mm/week movement.
The randomization list was generated by a customized software, allowing a random list with an allocation ratio 1:1.
All the statistical analysis was blindly performed in regards of patient’s group origin.
All the measurements were repeated by the same operator one month after the first examination and intraclass correlation coefficient was calculated both for the apex and canine tip. Intraclass Correlation Coefficient (ICC) values were 0.87 and 0.88 for canine tip and canine root apex respectively.
Descriptive statistics are expressed as median and interquartile ranges. The data were tested for normality using the Shapiro-Wilk test. The nonparametric Spearman’s rank correlation test was used to evaluate the dependence among the measured characteristics.
The nonparametric Mann-Whitney U test was used to evaluate differences between groups. Differences with a p-value less than 0.01 were selected as significant and data were acquired and analyzed using R v3.4.4 software environment .