Introduction
The objective of this study was to evaluate the in vivo transfer accuracy of a new slim and rigid computer-aided design and manufacturing (CAD-CAM) transfer tray.
Methods
Seventeen patients (9 males and 8 females; average age, 14.3 years) with permanent dentition undergoing orthodontic treatment were enrolled. Digital models were obtained, and brackets and molar tubes were virtually positioned. The CAD-CAM transfer tray was designed and 3-dimensionally printed, and the bonding procedure was performed. Transfer accuracy was assessed by measuring linear and angular deviations between planned and in vivo bonded bracket positions by superimposition using a local best-fit alignment performed by a semiautomatic algorithm in the analysis software Geomagic Control (3D System Inc, Rock Hill, NC).
Results
A total of 408 teeth (208 maxillary and 200 mandibular) were superimposed. All mean linear deviations were <0.5 mm, although some exceeded the 0.25-mm cutoff (4.4% of mesiodistal measurements, 4.2% of vertical, and 0.2% of buccolingual). Angular discrepancies were greater, especially when using a 1° cutoff (25.7% for torque, 33.8% for rotation, and 18.6% for tip). However, these discrepancies decreased significantly when the cutoff was raised to 2° (in 5.6%, 9.3% and 3.4% respectively). Considering all brackets, no directional bias was detected, except in buccolingual measurements (35% vestibular and 65% lingual). Transfer accuracy was influenced by the variables tooth type, tooth position, and arch, which had various impacts across specific measurements.
Conclusions
The new CAD-CAM transfer tray demonstrated good overall transfer accuracy and some clinical advantages, although refinement of the design is required to enhance its performance.
Highlights
-
•
To test the accuracy of a new, slim, and rigid CAD-CAM transfer tray in vivo.
-
•
This low-profile design presents several advantages for clinicians.
-
•
In vivo scan models and computer-aided design models are digitally superimposed tooth by tooth.
-
•
Linear and angular deviations are investigated for each bracket and tube.
-
•
The results are discussed and explained with reference to the existing literature.
In straightwire orthodontics, accurate bracket placement minimizes the need for bracket repositioning or archwire bends during finishing. The indirect bonding technique appears to offer some advantages over direct bonding. These include a reduction in chairside time, shorter treatment duration, decreased need for bracket repositioning, and improved occlusal outcomes. ,, This should lead to a reduction in overall cost, although Czolgosz et al disagree with this assertion; in contrast, some of the above-cited advantages have not been reported by all authors.
Indirect bonding seems to entail a higher incidence of accidental bonding failures, likely because of reduced moisture control and insufficient composite polymerization. , It may also cost more time and resources because of the laboratory work required. However, digital bonding somewhat mitigates this concern, as the digital workflow is more streamlined and, therefore, quicker than manual processes.
Transfer trays for indirect bonding can now be computer-aided design and manufacturing (CAD-CAM), together with virtual planning of bracket positions. The dedicated software enables previsualization of orthodontic outcomes and appliances to be customized at various levels, , thereby overcoming anatomic variations in vestibular tooth surfaces. However, trays must be accurate enough to replicate digitally planned bracket positions, as proved by several studies ,,,,,,, and 2 recent meta-analyses. , In fact, CAD-CAM transfer trays appear sufficiently accurate to meet the American Board of Orthodontics (ABO)-Objective Grading System (OGS) standards for bracket placement, although 0.25 mm and 1° have been used by some. ,,,,
Although polyvinyl siloxane transfer trays remain the gold standard, above both vacuum-formed and CAD-CAM transfer trays, , the transfer accuracy of the latter is clinically comparable, particularly if fabricated with soft materials. Although providing excellent precision by completely covering brackets and molar tubes, helping to stabilize them, such a design tends to hinder direct visualization of bracket positions and visual assessment of the contact between the tooth surface and the bracket base. Moreover, excess composite material cannot be removed before curing, resulting in an excessive composite flash around the bracket. When using a bulky transfer tray, effective light-curing can also be challenging because the intensity of light diminishes with the distance from the bracket base or molar tube, potentially explaining why the bracket failure rate is higher than in direct bonding. ,
To address and overcome these limitations, a new, slim, and rigid CAD-CAM was tested in vivo for transfer accuracy. The null hypothesis was that, despite its minimal bulk and reduced profile, the new transfer tray would achieve accuracy levels that comply with ABO-OGS requirements.
Material and methods
Seventeen patients (9 males and 8 females; average age, 14.3 ± 3.7 years) requiring vestibular fixed orthodontic treatment were consecutively enrolled in this prospective study at the University of Ferrara Department of Orthodontics. Patients had been previously treated or not with orthopedic or functional appliances, and none presented enamel or syndromic diseases. The index of orthodontic treatment need was calculated for each patient by a single expert orthodontist (L.L.), and patients with a score of >3 were enrolled. All enrolled patients had permanent dentition, although not all had all clinical tooth crowns available for bracket placement because of poor eruption ( Supplementary Table ). Therefore, a total of 452 statistical units were available for evaluation.
All patients, or guardians of young patients, gave consent to the clinical protocol for the study, which was approved by the University of Ferrara Postgraduate School of Orthodontics Ethics Committee as protocol number 11/2023.
After obtaining pretreatment orthodontic records (photographs and radiographs), stereolithography format (STL) digital models of the maxillary and mandibular arches were generated using a Carestream CS3600 intraoral scanner (Carestream Dental LLC, Atlanta, GA). Then, after the tooth segmentation process was executed using Orthoanalyzer 2021 software (3-Shape, Copenhagen, Denmark), the STL files were imported into Rhinoceros software (Robert McNeel & Associates, Seattle, Wash) for positioning each bracket and tube, adhering strictly to Andrew’s indications. The facial axis clinical crown lines were traced for each tooth and, subsequently, both brackets and molar tubes were automatically positioned using an algorithm created via Rhinoceros Grasshopper plugin (Robert McNeel & Associates, Seattle, WA), which automatically identified the facial axis midpoint on the facial axis clinical crown line drawn previously. The vertical position was thus checked and adjusted according to McLaughlin et al.
Preadjusted edgewise brackets with 0.022-in slots were virtually placed on incisors, canines, and premolars, and nonconvertible tubes with 0.022-in slots on molars (Primo; Sweden & Martina, Due Carrare, Italy). One operator (M.P.) performed all digital bonding procedures, which were checked and approved by a senior expert orthodontist (L.L.). The Rhinoceros Grasshopper plugin was then used to automatically and instantly design a reduced-bulk transfer tray to house the brackets and tubes in their virtually approved positions.
The transfer tray design featured a single core with arms protruding in a vestibular direction from each tooth ( Fig 1 , A ) for securely holding the brackets and molar tubes using different retention mechanisms ( Fig 1 , B and C ). The core was designed to touch the occlusal surfaces of the posterior teeth, partially covering both the vestibular and lingual coronal surfaces of the premolars and molars by 1-2 mm. It also covered the entire lingual surface of the canines and incisors, extending 1.5-2.0 mm from their cusp tips and incisal edges, respectively, onto their vestibular surface.
A, Digital design of the new, slim, rigid CAD-CAM transfer tray; B, Detail of the transfer tray arms with core; C, After core removal.
A dual holding mechanism was employed for all brackets, the first being based primarily on active clamping forces between the tray arm and the vertical bracket slots. This was ensured by the mesiodistal width of each arm, which was designed to be 0.05 mm wider than the vertical slot. This slight dimensional discrepancy required the bracket to be firmly wedged into place during positioning, creating a stabilizing clamping force ( Fig 2 , A – C ). The second mechanism aimed to stabilize the vertical position of the brackets through 2 rounded stops on each tray arm placed in contact with the upper and lower surfaces of the bracket wings ( Fig 2 , A – C ), as well as the upper and lower edges of the bracket base at its central extremities ( Fig 2 , D and E ). The upper stop could be described as a vertically flattened hemisphere with a radius ranging 0.8-1.0 mm, depending on the bracket type. Its position also varied depending on the tooth type: on the incisors and canines, the stops were more centered, whereas the vestibular side of the arm appeared rounded ( Fig 2 , B ). However, on the premolars, the stops were oriented internally and coronally, and the vestibular side of the arm appeared flat ( Fig 2 , C ). The lower stop could be described as an internally oriented sphere with a radius ranging 0.85-1.00 mm, depending on the bracket type ( Fig 2 , B and D).
Transfer tray arm with the respective bracket in position (A) ; with detail of incisor (B) and premolar brackets (C) from a lateral view; and from gingival (D) and occlusal (E) perspectives.
Transfer tray arm with molar tube in position (A), and view of an outer (B) and inner surface (C).
Each tray arm was designed with a cage mechanism to house the molar tubes ( Fig 3 , A ) featuring a vertically flattened semicircular-shaped hole of maximum (central) diameter 1 mm on the vestibular side to accommodate the molar tube hook ( Fig 3 , B ). The internal surface of the cage fit the molar tube profile, with a contour that ensured close contact with the vestibular surface of their bases. A vertical support, designed to touch the gingival surface of the molar tube, was incorporated into the design ( Fig 3 , C ).
The resulting transfer tray was printed using a Pro 95 digital light processing 3-dimensional (3D) printer (SprintRay Europe GmbH, Weiterstadt, Germany), with Z-axis resolution set to 100 μm. The resin was Keyprint KeySplint Soft (Keystone Industries, Singen, Germany), a biocompatible, Class II off-label certified, light violet, translucent material featuring high stiffness (Shore D hardness, 80-85) and moderate resistance to deformation (flexural modulus of 135-200 MPa) and breakage (elongation of 110% on breakage). After slicing, using RayWare software (SprintRay GmbH, Weiterstadt, Germany), the transfer trays were printed via digital light processing technology. The printing direction was vertical and oriented upward. The postcuring process involved cleaning with 95% isopropyl alcohol in a centrifuge for 10 minutes, drying with compressed air, and then curing for 5 minutes under a nitrogen-filled ultraviolet lamp (TeraHertz Cure, Graphy, Seoul, South Korea). Subsequently, the transfer tray was cut into 3 sections: 1 anterior (for the incisors and canines) and 2 posteriors (for the premolars and molars).
After that, both brackets and molar tubes were positioned in the CAD-CAM transfer tray. The brackets were clamped, and the molar tubes were secured and stabilized within the cage via a small amount of red wax, thereby creating a custom-fitted housing ( Fig 4 ).
Process of manufacturing the clear, slim, rigid, and CAD-CAM transfer tray.
The same operator (M.P.) performed indirect bonding in a single session, having ensured the fit and stability of each transfer tray in vivo. After isolating the operating field using the NOLA Dry Field System (Great Lakes Dental Technologies, Tonawanda, NY), each tooth was etched for 20 seconds. The teeth were then rinsed and dried, and a self-adhesive, low-viscosity orthodontic composite (GC Kommonbase HV Clear; GC Orthodontics Europe GmbH, Breckerfeld, Germany) was applied to the bracket and tube bases. Before this, the surfaces were cleaned with acetone using a microbrush and subsequently dried.
After the composite application, the transfer tray was positioned, and gentle finger pressure was used to ensure proper seating. The close contact between the bracket bases and crown surfaces was visually inspected, and excess composite was removed using a microbrush. Polymerization was then performed using a light-emitting diode light-curing lamp (ProHALO, Prodonto, Mainz-Hechtsheim, Germany) for 10 seconds per tooth in a sequential, tooth-by-tooth manner from the front (light emission intensity, 3000 mw/cm ).
After the transfer tray removal, any residual wax on the molar tubes and excess composite beneath the brackets (the only area inaccessible to the microbrush) were removed. In instances of bonding failure, bracket repositioning was not feasible because of the distortion of the rigid CAD-CAM transfer tray at the arms, and so was achieved via direct bonding, excluding such brackets from the subsequent analysis.
Finally, a thin layer of mattifying blue powder (Dreve Dentamid GmbH, Unna, Germany) was applied to the surfaces of the brackets and molar tubes, after compressed air drying for 5 seconds. A second digital impression of both arches was obtained, and the resulting STL files were used as in vivo models to investigate the accuracy of the transfer tray via the digital workflow described.
Rhinoceros software was used to export the respective CAD malocclusion models, which included the digitally planned positions of brackets and molar tubes, into STL format to serve as a reference for subsequent linear and angular measurements. The optimized in vivo and CAD models ( Fig 5 ) were created and digitally superimposed on each tooth and positional bracket, and molar tube discrepancies were measured using Geomagic Control software (3D Systems Inc, Rock Hill, NC) following a local, best-fit alignment for each tooth and employing a semiautomated algorithm to improve efficiency and standardize measurements (Supplementary Fig). Further details on both procedures can be found in studies by Koch et al and Palone et al.
Digital steps for creating the optimized scanned in vivo model and the corresponding optimized CAD model before digital superimposition.
Statistical analysis
Any invalid data points because of scanning deficiencies (particularly low-quality scans of the bracket and molar tube profiles), which hindered superimposition of CAD patches and therefore generation of optimized in vivo models, were flagged and excluded from subsequent analysis. The percentages of these invalid measurements were calculated, and data discrepancies are reported in both absolute and directional (positive or negative) terms, summarizing their clinical significance based on the mouth quadrant and measurement type ( Table I ).
Table I
Meaning of directional bias for positive and negative values by quadrant
| Variables | Quadrant | |||
|---|---|---|---|---|
| 1 | 2 | 3 | 4 | |
| x-Linear | ||||
| Negative | Distal | Mesial | Mesial | Distal |
| Positive | Mesial | Distal | Distal | Mesial |
| y-Linear | ||||
| Negative | Occlusal | Occlusal | Gingival | Gingival |
| Positive | Gingival | Gingival | Occlusal | Occlusal |
| z-Linear | ||||
| Negative | Lingual | Lingual | Lingual | Lingual |
| Positive | Vestibular | Vestibular | Vestibular | Vestibular |
| x-Rotation | ||||
| Negative | Negative torque | Negative torque | Positive torque | Positive torque |
| Positive | Positive torque | Positive torque | Negative torque | Negative torque |
| y-Rotation | ||||
| Negative | Mesiorotation | Distorotation | Distorotation | Mesiorotation |
| Positive | Distorotation | Mesiorotation | Mesiorotation | Distorotation |
| z-Rotation | ||||
| Negative | Positive tip | Negative tip | Negative tip | Positive tip |
| Positive | Negative tip | Positive tip | Positive tip | Negative tip |
Descriptive analysis was performed, calculating the mean absolute deviation of both linear and angular measurements by tooth type, single arch, and for both arches. Deviation thresholds of 0.25 mm for linear and 1° for angular measurements were set at half of those recommended by the ABO (0.5 mm and 2°), considering the worst-case scenario of incorrectly placing brackets in opposite directions on 2 adjacent teeth, potentially compromising the marginal ridge leveling by 0.5 mm. Nonetheless, the ABO threshold was also considered to enable clinical comparison with other studies. The percentage of directional errors was also calculated.
To predict outcomes, classification and regression tree (CART) analysis, rather than inferential methods, was conducted using R Software; CART can handle interactions between multiple variables (tooth, tooth type, and arch) and reveal the hierarchical relationships between predictors and outcomes. The proposed splits were statistically significant at the significance threshold set at 5%. Because CART models are complex and nonlinear, there is no definitive method for determining the required sample size for a given power in advance. However, judging by similar studies analyzing around ≤200 brackets, , the sample size in this study was appropriate.
Results
Eighteen immediate bonding failures were recorded and excluded from subsequent analysis (3.98% of the total), which was therefore conducted on a total of 434 observations. However, 26 invalid measurements (5.46% of the total), because of poor in vivo scan quality (despite the use of blue mattifying powder), were also excluded. Hence, 208 maxillary brackets and 200 mandibular brackets were evaluated on a total of 408 superimpositions. Considering 3 linear and 3 angular measurements, a total of 2448 measurements was obtained.
Bonding discrepancies in both linear and angular measurements, considering tooth type, single arch, and both arches, were calculated, with none exceeding the 0.25 mm and 1° thresholds, except for rotation discrepancies at the incisors (1.007° ± 0.810°) and canines (1.002° ± 0.794°). When considering tooth types, the average discrepancy ranged from a minimum of 0.031 ± 0.027 mm at the incisor (buccolingual) to a maximum of 0.104 ± 0.096 mm at the molar (vertical). For angular values, discrepancies ranged from a minimum of 0.339° ± 0.354° at the molar (tip) to a maximum of 1.007° ± 0.810° at the incisor (rotation) ( Table II ).
Table II
Linear and angular measurement by tooth type, single arch, and in both arches
| Groups | n | Mesiodistal (mm) | Vertical (mm) | Buccolingual (mm) | Torque (°) | Rotation (°) | Tip (°) |
|---|---|---|---|---|---|---|---|
| Incisor | 124 | 0.095 ± 0.066 | 0.064 ± 0.052 | 0.031 ± 0.027 | 0.552 ± 0.573 | 1.007 ± 0.810 | 0.772 ± 0.556 |
| Canine | 65 | 0.094 ± 0.076 | 0.076 ± 0.063 | 0.036 ± 0.042 | 0.837 ± 0.684 | 1.002 ± 0.794 | 0.737 ± 0.594 |
| Premolar | 118 | 0.092 ± 0.075 | 0.080 ± 0.067 | 0.040 ± 0.036 | 0.780 ± 0.674 | 0.935 ± 0.728 | 0.666 ± 0.628 |
| Molar | 101 | 0.095 ± 0.090 | 0.104 ± 0.096 | 0.045 ± 0.035 | 0.848 ± 0.764 | 0.699 ± 0.602 | 0.339 ± 0.354 |
| Maxilla | 208 | 0.093 ± 0.078 | 0.073 ± 0.068 | 0.035 ± 0.033 | 0.672 ± 0.646 | 0.866 ± 0.687 | 0.589 ± 0.464 |
| Mandible | 200 | 0.094 ± 0.076 | 0.088 ± 0.075 | 0.041 ± 0.037 | 0.804 ± 0.709 | 0.954 ± 0.799 | 0.670 ± 0.658 |
| Both arches | 408 | 0.094 ± 0.077 | 0.080 ± 0.072 | 0.038 ± 0.035 | 0.737 ± 0.680 | 0.909 ± 0.745 | 0.629 ± 0.568 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
