Potentialities and limitations of computer-aided design and manufacturing technology in the nonextraction treatment of Class I malocclusion


Computer-aided design and manufacturing (CAD-CAM) systems have assisted orthodontists to position brackets virtually. The purpose of this study was to evaluate if a CAD-CAM system could predict the orthodontic treatment outcome of patients with Angle Class I malocclusion with mild crowding or spacing and with no need for orthodontic extraction.


Using the American Board of Orthodontics Cast-Radiograph Evaluation (ABO-CRE) and color map superimposition, the treated occlusion was compared with the virtual final occlusion of 24 young adults with Class I occlusion. Using eXceed software (eXceed, Witten, Germany), we created the final occlusion prediction for each patient (virtual set up group). A digital model of the final occlusion of each patient was created (treated occlusion group). ABO-CRE score was used to compare groups. In addition, a color map was created for all subjects to access the mean and range values between the virtual set up model and treated occlusion model of each patient. Random and systematic errors were calculated. In addition, chi-square and t test were used.


Comparisons between virtual set up occlusion and treated occlusion showed statistically significant differences in 3 out of 7 measurements: interproximal contact score was larger for treated than virtual occlusion (0.45 mm and 0.04 mm, respectively), and the treated occlusion showed larger values than the virtual occlusion for occlusal contacts (14.13 mm and 7.62 mm, respectively) and overjet (7.37 mm and 0.66 mm, respectively). Although the treated occlusion showed a larger score than the virtual occlusion (50.41 mm and 34.58 mm, respectively), there is no significant difference between both. Root angulation decreased (from 1.95 ± 1.29 to 0.65 ± 0.71) because of the treatment.


ABO-CRE overall score presents no difference between groups. In addition, CAD-CAM setup occlusion closely predicts the final teeth alignment and leveling with interarch relationships showing less ABO-CRE score deduction.


  • CAD/CAM technology can closely predict final occlusion treatments.

  • CAD/CAM technology allows accurate bracket placement and improved root angulation.

  • Virtual and treated occlusions show similar American Board of Orthodontics score deductions.

  • Interarch evaluation had more ABO-CRE deductions for treated occlusion than virtual occlusion.

While achieving pleasing esthetic treatment in the shortest time possible, the orthodontic treatment should be both effective and efficient, leading to fewer appointments for orthodontists and patients. , Ideal bracket placement allows a high-quality treatment with reduced number of consults and chair time. Therefore, accurate bracket placement at the bonding appointment is still critically important for both the orthodontist and the patient.

Computer-aided design and manufacturing (CAD-CAM) systems have helped the orthodontist precisely place brackets to minimize the need for repositioning of brackets and bends in the archwires. An important innovation of this system is the customization of orthodontic brackets; that is, adding pads to adjust for differences in dental anatomy and/or tooth size. Technology has greatly affected orthodontists, and through use of this knowledge has improved the reproducibility, efficiency, and enhanced treatment.

The CAD-CAM technology has stood out in the market for developing solutions directed toward planning, positioning, and customizing brackets virtually. This system aims to make the orthodontic office more efficient, simplifying processes and assisting clinicians and laboratories in manufacturing accurate devices. To date, no study has evaluated the eXceed (Witten, Germany) system related to orthodontic treatment regarding bracket positioning and its respective dental alveolar changes.

The purpose of this study was to evaluate, using the Cast-Radiograph Evaluation (CRE) system of the American Board of Orthodontics (ABO), whether CAD-CAM system setup can be used to predict orthodontic treatment results of patients with Class I malocclusion with mild crowding or spacing and not requiring orthodontic extraction. The null hypothesis was that there were no differences between the CAD-CAM system ideal setup measurements and the full postalignment and leveling orthodontic treatment.

Material and methods

The samples consisted of 24 Angle Class I malocclusion young adults aged over 18 years (15 women, 9 men; mean age, 26.14 ± 6.53 years). All patients were treated in accordance with the Research Ethics Committee under protocol number 2.451.252 of the São Paulo State University Araraquara, São Paulo, Brazil.

The requirements for the inclusion of clinical subjects were as follows: (1) complete permanent dentition (excluding third molars); (2) molar relationship of Angle Class I on both sides of the dental arch; (3) canine relationship of Angle Class I or maximum one fourth Class II; (4) none or slight dental posterior crossbite; and (5) good overall health and normal periodontal condition.

The exclusion criteria were as follows: (1) caries lesions that compromise the dental structure; (2) morphologic variation in size and shape of the crown; (3) moderate open bite or deep bite equal or greater than 4 mm; and (4) crowding or tooth spacing greater than 4 mm.

Computed tomographic images were obtained within 30 days before bracket bonding and immediately after debonding. The cone-beam computed tomography scans were acquired with an i-CAT (KaVo Dental GmbH, Germany) unit at these settings: 3.7 mA, 120 kV, exposure time of 40 seconds, voxel size of 0.2 mm, axial slice thickness of 0.3 mm, and scanning area of 20 cm x 25 cm. OnDemand3D (version; Cybermed Inc, Seoul, Korea) was used for multiplanar reconstruction and measurements. The Dental Volume Reformat module allowed the generation of the panoramic radiograph, which was used to control, diagnose, and evaluate the parallelism between the roots of the ABO root angulation criteria.

Two groups of 24 digital 3-dimensional (3D) digital models each were evaluated: eXceed virtual setup (group 1) and treated occlusion (group 2). Three-dimensional digital models were obtained according to the following steps:

  • 1.

    Traditional plaster models were obtained with alginate impressions of the dental arches using Jeltrate Plus (Dentsply Sirona, Rio de Janeiro, Brazil), followed with Durone plaster type IV (Dentsply Sirona). The next step was acquiring 3D models from the plaster casts of the patients by digitization using an R700 desktop scanner (3Shape, Copenhagen, Denmark). The reconstruction of the digital models was obtained by ScanIt software (3Shape), producing files in the type 3Shape zip format (.3sz). The files were exported to Ortho Analyzer 2013 software (Kulzer, Hanau, Germany) for conversion to stereolithography (.stl file format).

  • 2.

    All data of the 24 digital models were exported to platform eXceed to create the virtual setups carried out by Doctor WebGL Software 2.0 (eXceed) analysis and diagnostic tools. EasyClip Plus self-ligating brackets 0.022 × 0.028-in slot (Aditek Orthodontics, Cravinhos, Brazil) were used for all patients. The treatment planning was checked by the company’s staff, revised, and approved by only 1 operator (F.C.M).

  • 3.

    The system generated 3D models of all patients to be printed by an Objet Eden500V 3D Printer (Stratasys Ltd, Valencia, Calif) using VeroDent photopolymer (Stratasys Ltd). Indirect bond devices were manufactured by the thermoforming method using Biolon and Drufolen H (Dreve Dentamid GmbH, Unna, Germany), each overlapped with a transparent, flexible, 1-mm thick ethylene-vinyl acetate coated with another polyethylene terephthalate, which is more rigid to prevent distortion during an indirect bonding procedure.

The indirect bonding method was performed in all 24 patients sample according to the following sequence technique:

  • 1.

    Enamel etching technique was performed on all dental arch teeth for 20 seconds, followed by washing and drying of the enamel. Transbond XT primer (3M Unitek, Monrovia, Calif) was applied on both bracket base and conditioned enamel surface, followed by enamel polymerization for 20 seconds. First, the tray was positioned on the maxillary arch, followed by 20 seconds of polymerization on each bracket side. The same procedures were performed at the mandibular arch. After brackets, adhesive curing the double layer trays was removed, first the rigid plate followed by the silicone plate (more flexible), both in the maxillary arch. The same procedures were performed in the mandibular arch.

  • 2.

    Damon archwires (Ormco, Orange, Calif) were used for alignment and leveling, according to the following sequence: 0.014, 0.018, 0.017 × 0.025-in, 0.019 × 0.025-in nickel-titanium, and 0.019 × 0.025-in stainless steel coordinates on the diagram sent by the eXceed system. No inter- or intramaxillary elastic and/or wire bending was implemented during treatment to avoid interference with the ideal planning programmed by eXceed.

  • 3.

    After 3 months using rectangular steel archwire 0.019 × 0.025-in, digital models were obtained following the same initial methodology previously described. Data samples were collected, and all patients were treated according to the previous treatment plan and biomechanics.

Virtual setup and treated occlusion 3D models’ files were exported into Geomagic Design X (3D Systems, Rock Hill, SC) to set landmarks on each tooth to perform measurements (scale of 1/100 mm). ABO linear measurements were obtained by measuring the distance between 2 adjacent vectors or landmarks to vector created on the tooth according to each analyzed criteria ( Fig 1 ). The measurements were obtained according to the parallelism of the global axes to maintain standardization and avoid any bias.

Fig 1
Demarcated points on the teeth for ABO measurements. Alignment and marginal ridges were measured using 2 adjacent vectors (A) ; buccolingual inclination measurements were obtained by creating a transversal axis between both opposite teeth (B) . In the sagittal view, landmarks to overjet (C) , occlusal relationship (D) , and occlusal contacts (E) .

Before measurements, 1 observer (F.C.M) was trained and calibrated to the ABO-CRE and accomplished overall 3D model analyses. Virtual setup and treated occlusion were objectively quantified and compared according to the ABO-CRE: alignment and rotations, marginal ridges (MR), buccolingual inclination, occlusal relationship, occlusal contact, overjet, and interproximal contact. Root parallelism was evaluated only on the treated occlusion group.

The superimposition method was carried out to compare the virtual setup and treated occlusion 3D digital models using Geomagic Control X. After measurements, all models were trimmed at tooth edge and aligned using the best-fit tool to compare geometric shape surface. A color map was created for all subjects to access the mean and ranging values between the “cloud points” of the 2 digital group models.

Statistical analysis

Data analyses were performed by Microsoft Excel 2013 (Microsoft Corp, Redmond, Wash) and GraphPad Prism (version 7; GraphPad Software, San Diego, Calif). The homogeneity of variance and normality of the residuals were confirmed by the Levene test and Shapiro-Wilk and D’Agostino and Pearson normality tests, respectively. Intraobserver systematic errors between the replicate were described as mean differences and compared statistically with paired t tests. Intraobserver random error was estimated using intraclass correlation coefficients (ICCs) and method errors [√ (∑d 2 /2n)]. In addition, Bland-Altman analysis was carried out to verify the agreement between all measurement criteria. Descriptive statistics and chi-square analyses were performed for each ABO-CRE category to verify the percentage of linear measurements that were in agreement within acceptable limits. Student t test and Pearson correlation coefficient analyzed the superimposition of the two 3D model groups. Clinical significance for teeth movements followed the ABO guidelines of >0.05 mm for linear movements. All statistical tests were performed at a significance level of 5%.


Intraobserver systematic errors of the 2 methods were similar ( Table I ). Method errors ranged from <0.01 mm to 0.23 mm ( Table II ). No pattern could be observed between both virtual setup and treated occlusion. ICCs, ranging from 0.658 to >0.999, showed a moderately high and high degree of reliability ( Table II ). Bland-Altman plots showed that the differences of repeatability between 2 methods were within the limits of agreement ( Fig 2 ).

Table I
Systematic differences (mm) between virtual setup measures and treated occlusion subjects
Variables Virtual setup Treated occlusion
Difference (mm) Sig Difference (mm) Sig
Alignment −0.04 0.517 −0.03 0.663
Marginal ridges 0.06 0.198 −0.03 0.467
Interproximal contacts 0.00 N/A 0.00 0.327
Buccolingual inclination −0.04 0.441 −0.15 0.357
Occlusal relationship −0.01 0.872 −0.04 0.464
Occlusal contacts 0.02 0.519 −0.09 0.110
Overjet 0.01 0.737 −0.08 0.374

Note. Significance level at P <0.05.
Sig , significance; N/A , not applicable.

Jan 30, 2021 | Posted by in Orthodontics | Comments Off on Potentialities and limitations of computer-aided design and manufacturing technology in the nonextraction treatment of Class I malocclusion
Premium Wordpress Themes by UFO Themes